Datasets:

CShorten

/

Last-Week-on-ML-ArXiv

like 0

Cascading Failures in Smart Grids under Random, Targeted and Adaptive Attacks

We study cascading failures in smart grids, where an attacker selectively compromises the nodes with probabilities proportional to their degrees, betweenness, or clustering coefficient. This implies that nodes with high degrees, betweenness, or clustering coefficients are attacked with higher probability. We mathematically and experimentally analyze the sizes of the giant components of the networks under different types of targeted attacks, and compare the results with the corresponding sizes under random attacks. We show that networks disintegrate faster for targeted attacks compared to random attacks. A targeted attack on a small fraction of high degree nodes disintegrates one or both of the networks, whereas both the networks contain giant components for random attack on the same fraction of nodes. An important observation is that an attacker has an advantage if it compromises nodes based on their betweenness, rather than based on degree or clustering coefficient. We next study adaptive attacks, where an attacker compromises nodes in rounds. Here, some nodes are compromised in each round based on their degree, betweenness or clustering coefficients, instead of compromising all nodes together. In this case, the degree, betweenness, or clustering coefficient is calculated before the start of each round, instead of at the beginning. We show experimentally that an adversary has an advantage in this adaptive approach, compared to compromising the same number of nodes all at once.

Aligning Artificial Intelligence with Humans through Public Policy

Given that Artificial Intelligence (AI) increasingly permeates our lives, it is critical that we systematically align AI objectives with the goals and values of humans. The human-AI alignment problem stems from the impracticality of explicitly specifying the rewards that AI models should receive for all the actions they could take in all relevant states of the world. One possible solution, then, is to leverage the capabilities of AI models to learn those rewards implicitly from a rich source of data describing human values in a wide range of contexts. The democratic policy-making process produces just such data by developing specific rules, flexible standards, interpretable guidelines, and generalizable precedents that synthesize citizens' preferences over potential actions taken in many states of the world. Therefore, computationally encoding public policies to make them legible to AI systems should be an important part of a socio-technical approach to the broader human-AI alignment puzzle. This Essay outlines research on AI that learn structures in policy data that can be leveraged for downstream tasks. As a demonstration of the ability of AI to comprehend policy, we provide a case study of an AI system that predicts the relevance of proposed legislation to any given publicly traded company and its likely effect on that company. We believe this represents the "comprehension" phase of AI and policy, but leveraging policy as a key source of human values to align AI requires "understanding" policy. Solving the alignment problem is crucial to ensuring that AI is beneficial both individually (to the person or group deploying the AI) and socially. As AI systems are given increasing responsibility in high-stakes contexts, integrating democratically-determined policy into those systems could align their behavior with human goals in a way that is responsive to a constantly evolving society.

Cactus Mechanisms: Optimal Differential Privacy Mechanisms in the Large-Composition Regime

Most differential privacy mechanisms are applied (i.e., composed) numerous times on sensitive data. We study the design of optimal differential privacy mechanisms in the limit of a large number of compositions. As a consequence of the law of large numbers, in this regime the best privacy mechanism is the one that minimizes the Kullback-Leibler divergence between the conditional output distributions of the mechanism given two different inputs. We formulate an optimization problem to minimize this divergence subject to a cost constraint on the noise. We first prove that additive mechanisms are optimal. Since the optimization problem is infinite dimensional, it cannot be solved directly; nevertheless, we quantize the problem to derive near-optimal additive mechanisms that we call "cactus mechanisms" due to their shape. We show that our quantization approach can be arbitrarily close to an optimal mechanism. Surprisingly, for quadratic cost, the Gaussian mechanism is strictly sub-optimal compared to this cactus mechanism. Finally, we provide numerical results which indicate that cactus mechanism outperforms the Gaussian mechanism for a finite number of compositions.

Defending Multimodal Fusion Models against Single-Source Adversaries

Beyond achieving high performance across many vision tasks, multimodal models are expected to be robust to single-source faults due to the availability of redundant information between modalities. In this paper, we investigate the robustness of multimodal neural networks against worst-case (i.e., adversarial) perturbations on a single modality. We first show that standard multimodal fusion models are vulnerable to single-source adversaries: an attack on any single modality can overcome the correct information from multiple unperturbed modalities and cause the model to fail. This surprising vulnerability holds across diverse multimodal tasks and necessitates a solution. Motivated by this finding, we propose an adversarially robust fusion strategy that trains the model to compare information coming from all the input sources, detect inconsistencies in the perturbed modality compared to the other modalities, and only allow information from the unperturbed modalities to pass through. Our approach significantly improves on state-of-the-art methods in single-source robustness, achieving gains of 7.8-25.2% on action recognition, 19.7-48.2% on object detection, and 1.6-6.7% on sentiment analysis, without degrading performance on unperturbed (i.e., clean) data.

Protoformer: Embedding Prototypes for Transformers

Transformers have been widely applied in text classification. Unfortunately, real-world data contain anomalies and noisy labels that cause challenges for state-of-art Transformers. This paper proposes Protoformer, a novel self-learning framework for Transformers that can leverage problematic samples for text classification. Protoformer features a selection mechanism for embedding samples that allows us to efficiently extract and utilize anomalies prototypes and difficult class prototypes. We demonstrated such capabilities on datasets with diverse textual structures (e.g., Twitter, IMDB, ArXiv). We also applied the framework to several models. The results indicate that Protoformer can improve current Transformers in various empirical settings.

Bayesian Optimization Over Iterative Learners with Structured Responses: A Budget-aware Planning Approach

The rising growth of deep neural networks (DNNs) and datasets in size motivates the need for efficient solutions for simultaneous model selection and training. Many methods for hyperparameter optimization (HPO) of iterative learners including DNNs attempt to solve this problem by querying and learning a response surface while searching for the optimum of that surface. However, many of these methods make myopic queries, do not consider prior knowledge about the response structure, and/or perform biased cost-aware search, all of which exacerbate identifying the best-performing model when a total cost budget is specified. This paper proposes a novel approach referred to as Budget-Aware Planning for Iterative Learners (BAPI) to solve HPO problems under a constrained cost budget. BAPI is an efficient non-myopic Bayesian optimization solution that accounts for the budget and leverages the prior knowledge about the objective function and cost function to select better configurations and to take more informed decisions during the evaluation (training). Experiments on diverse HPO benchmarks for iterative learners show that BAPI performs better than state-of-the-art baselines in most of the cases.

Binary and Multinomial Classification through Evolutionary Symbolic Regression

We present three evolutionary symbolic regression-based classification algorithms for binary and multinomial datasets: GPLearnClf, CartesianClf, and ClaSyCo. Tested over 162 datasets and compared to three state-of-the-art machine learning algorithms -- XGBoost, LightGBM, and a deep neural network -- we find our algorithms to be competitive. Further, we demonstrate how to find the best method for one's dataset automatically, through the use of a state-of-the-art hyperparameter optimizer.

p-Meta: Towards On-device Deep Model Adaptation

Data collected by IoT devices are often private and have a large diversity across users. Therefore, learning requires pre-training a model with available representative data samples, deploying the pre-trained model on IoT devices, and adapting the deployed model on the device with local data. Such an on-device adaption for deep learning empowered applications demands data and memory efficiency. However, existing gradient-based meta learning schemes fail to support memory-efficient adaptation. To this end, we propose p-Meta, a new meta learning method that enforces structure-wise partial parameter updates while ensuring fast generalization to unseen tasks. Evaluations on few-shot image classification and reinforcement learning tasks show that p-Meta not only improves the accuracy but also substantially reduces the peak dynamic memory by a factor of 2.5 on average compared to state-of-the-art few-shot adaptation methods.

Anatomy-Guided Weakly-Supervised Abnormality Localization in Chest X-rays

Creating a large-scale dataset of abnormality annotation on medical images is a labor-intensive and costly task. Leveraging weak supervision from readily available data such as radiology reports can compensate lack of large-scale data for anomaly detection methods. However, most of the current methods only use image-level pathological observations, failing to utilize the relevant anatomy mentions in reports. Furthermore, Natural Language Processing (NLP)-mined weak labels are noisy due to label sparsity and linguistic ambiguity. We propose an Anatomy-Guided chest X-ray Network (AGXNet) to address these issues of weak annotation. Our framework consists of a cascade of two networks, one responsible for identifying anatomical abnormalities and the second responsible for pathological observations. The critical component in our framework is an anatomy-guided attention module that aids the downstream observation network in focusing on the relevant anatomical regions generated by the anatomy network. We use Positive Unlabeled (PU) learning to account for the fact that lack of mention does not necessarily mean a negative label. Our quantitative and qualitative results on the MIMIC-CXR dataset demonstrate the effectiveness of AGXNet in disease and anatomical abnormality localization. Experiments on the NIH Chest X-ray dataset show that the learned feature representations are transferable and can achieve the state-of-the-art performances in disease classification and competitive disease localization results. Our code is available at https://github.com/batmanlab/AGXNet

Data Augmentation techniques in time series domain: A survey and taxonomy

With the latest advances in deep learning generative models, it has not taken long to take advantage of their remarkable performance in the area of time series. Deep neural networks used to work with time series depend heavily on the breadth and consistency of the datasets used in training. These types of characteristic are not usually abundant in the real world, where they are usually limited and often with privacy constraints that must be guaranteed. Therefore, an effective way is to increase the number of data using \gls{da} techniques, either by adding noise or permutations and by generating new synthetic data. It is systematically review the current state-of-the-art in the area to provide an overview of all available algorithms and proposes a taxonomy of the most relevant researches. The efficiency of the different variants will be evaluated; as a vital part of the process, the different metrics to evaluate the performance and the main problems concerning each model will be analysed. The ultimate goal of this study is to provide a summary of the evolution and performance of areas that produce better results to guide future researchers in this field.

Improving Data-driven Heterogeneous Treatment Effect Estimation Under Structure Uncertainty

Estimating how a treatment affects units individually, known as heterogeneous treatment effect (HTE) estimation, is an essential part of decision-making and policy implementation. The accumulation of large amounts of data in many domains, such as healthcare and e-commerce, has led to increased interest in developing data-driven algorithms for estimating heterogeneous effects from observational and experimental data. However, these methods often make strong assumptions about the observed features and ignore the underlying causal model structure, which can lead to biased HTE estimation. At the same time, accounting for the causal structure of real-world data is rarely trivial since the causal mechanisms that gave rise to the data are typically unknown. To address this problem, we develop a feature selection method that considers each feature's value for HTE estimation and learns the relevant parts of the causal structure from data. We provide strong empirical evidence that our method improves existing data-driven HTE estimation methods under arbitrary underlying causal structures. Our results on synthetic, semi-synthetic, and real-world datasets show that our feature selection algorithm leads to lower HTE estimation error.

Defense against adversarial attacks on deep convolutional neural networks through nonlocal denoising

Despite substantial advances in network architecture performance, the susceptibility of adversarial attacks makes deep learning challenging to implement in safety-critical applications. This paper proposes a data-centric approach to addressing this problem. A nonlocal denoising method with different luminance values has been used to generate adversarial examples from the Modified National Institute of Standards and Technology database (MNIST) and Canadian Institute for Advanced Research (CIFAR-10) data sets. Under perturbation, the method provided absolute accuracy improvements of up to 9.3% in the MNIST data set and 13% in the CIFAR-10 data set. Training using transformed images with higher luminance values increases the robustness of the classifier. We have shown that transfer learning is disadvantageous for adversarial machine learning. The results indicate that simple adversarial examples can improve resilience and make deep learning easier to apply in various applications.

Minority Report: A Graph Network Oracle for In Situ Visualization

In situ visualization techniques are hampered by a lack of foresight: crucial simulation phenomena can be missed due to a poor sampling rate or insufficient detail at critical timesteps. Keeping a human in the loop is impractical, and defining statistical triggers can be difficult. This paper demonstrates the potential for using a machine-learning-based simulation surrogate as an oracle to identify expected critical regions of a large-scale simulation. These critical regions are used to drive the in situ analysis, providing greater data fidelity and analysis resolution with an equivalent I/O budget to a traditional in situ framework. We develop a distributed asynchronous in situ visualization by integrating TACC Galaxy with CB-Geo MPM for material point simulation of granular flows. We employ a PyTorch-based 3D Graph Network Simulator (GNS) trained on granular flow problems as an oracle to predict the dynamics of granular flows. Critical regions of interests are manually tagged in GNS for in situ rendering in MPM.

Topology-aware Generalization of Decentralized SGD

This paper studies the algorithmic stability and generalizability of decentralized stochastic gradient descent (D-SGD). We prove that the consensus model learned by D-SGD is $\mathcal{O}{(m/N+1/m+\lambda^2)}$-stable in expectation in the non-convex non-smooth setting, where $N$ is the total sample size of the whole system, $m$ is the worker number, and $1-\lambda$ is the spectral gap that measures the connectivity of the communication topology. These results then deliver an $\mathcal{O}{(1/N+{({(m^{-1}\lambda^2)}^{\frac{\alpha}{2}}+ m^{-\alpha})}/{N^{1-\frac{\alpha}{2}}})}$ in-average generalization bound, which is non-vacuous even when $\lambda$ is closed to $1$, in contrast to vacuous as suggested by existing literature on the projected version of D-SGD. Our theory indicates that the generalizability of D-SGD has a positive correlation with the spectral gap, and can explain why consensus control in initial training phase can ensure better generalization. Experiments of VGG-11 and ResNet-18 on CIFAR-10, CIFAR-100 and Tiny-ImageNet justify our theory. To our best knowledge, this is the first work on the topology-aware generalization of vanilla D-SGD. Code is available at https://github.com/Raiden-Zhu/Generalization-of-DSGD.

Envelope imbalanced ensemble model with deep sample learning and local-global structure consistency

The class imbalance problem is important and challenging. Ensemble approaches are widely used to tackle this problem because of their effectiveness. However, existing ensemble methods are always applied into original samples, while not considering the structure information among original samples. The limitation will prevent the imbalanced learning from being better. Besides, research shows that the structure information among samples includes local and global structure information. Based on the analysis above, an imbalanced ensemble algorithm with the deep sample pre-envelope network (DSEN) and local-global structure consistency mechanism (LGSCM) is proposed here to solve the problem.This algorithm can guarantee high-quality deep envelope samples for considering the local manifold and global structures information, which is helpful for imbalance learning. First, the deep sample envelope pre-network (DSEN) is designed to mine structure information among samples.Then, the local manifold structure metric (LMSM) and global structure distribution metric (GSDM) are designed to construct LGSCM to enhance distribution consistency of interlayer samples. Next, the DSEN and LGSCM are put together to form the final deep sample envelope network (DSEN-LG). After that, base classifiers are applied on the layers of deep samples respectively.Finally, the predictive results from base classifiers are fused through bagging ensemble learning mechanism. To demonstrate the effectiveness of the proposed method, forty-four public datasets and more than ten representative relevant algorithms are chosen for verification. The experimental results show that the algorithm is significantly better than other imbalanced ensemble algorithms.

Learning to Infer 3D Shape Programs with Differentiable Renderer

Given everyday artifacts, such as tables and chairs, humans recognize high-level regularities within them, such as the symmetries of a table, the repetition of its legs, while possessing low-level priors of their geometries, e.g., surfaces are smooth and edges are sharp. This kind of knowledge constitutes an important part of human perceptual understanding and reasoning. Representations of and how to reason in such knowledge, and the acquisition thereof, are still open questions in artificial intelligence (AI) and cognitive science. Building on the previous proposal of the \emph{3D shape programs} representation alone with the accompanying neural generator and executor from \citet{tian2019learning}, we propose an analytical yet differentiable executor that is more faithful and controllable in interpreting shape programs (particularly in extrapolation) and more sample efficient (requires no training). These facilitate the generator's learning when ground truth programs are not available, and should be especially useful when new shape-program components are enrolled either by human designers or -- in the context of library learning -- algorithms themselves. Preliminary experiments on using it for adaptation illustrate the aforesaid advantages of the proposed module, encouraging similar methods being explored in building machines that learn to reason with the kind of knowledge described above, and even learn this knowledge itself.

Guided Exploration in Reinforcement Learning via Monte Carlo Critic Optimization

The class of deep deterministic off-policy algorithms is effectively applied to solve challenging continuous control problems. However, current approaches use random noise as a common exploration method that has several weaknesses, such as a need for manual adjusting on a given task and the absence of exploratory calibration during the training process. We address these challenges by proposing a novel guided exploration method that uses a differential directional controller to incorporate scalable exploratory action correction. An ensemble of Monte Carlo Critics that provides exploratory direction is presented as a controller. The proposed method improves the traditional exploration scheme by changing exploration dynamically. We then present a novel algorithm exploiting the proposed directional controller for both policy and critic modification. The presented algorithm outperforms modern reinforcement learning algorithms across a variety of problems from DMControl suite.

Trace Recovery from Stochastically Known Logs

In this work we propose an algorithm for trace recovery from stochastically known logs, a setting that is becoming more common with the increasing number of sensors and predictive models that generate uncertain data. The suggested approach calculates the conformance between a process model and a stochastically known trace and recovers the best alignment within this stochastic trace as the true trace. The paper offers an analysis of the impact of various cost models on trace recovery accuracy and makes use of a product multi-graph to compare alternative trace recovery options. The average accuracy of our approach, evaluated using two publicly available datasets, is impressive, with an average recovery accuracy score of 90-97%, significantly improving a common heuristic that chooses the most likely value for each uncertain activity. We believe that the effectiveness of the proposed algorithm in recovering correct traces from stochastically known logs may be a powerful aid for developing credible decision-making tools in uncertain settings.

Evaluation of Semantic Answer Similarity Metrics

There are several issues with the existing general machine translation or natural language generation evaluation metrics, and question-answering (QA) systems are indifferent in that context. To build robust QA systems, we need the ability to have equivalently robust evaluation systems to verify whether model predictions to questions are similar to ground-truth annotations. The ability to compare similarity based on semantics as opposed to pure string overlap is important to compare models fairly and to indicate more realistic acceptance criteria in real-life applications. We build upon the first to our knowledge paper that uses transformer-based model metrics to assess semantic answer similarity and achieve higher correlations to human judgement in the case of no lexical overlap. We propose cross-encoder augmented bi-encoder and BERTScore models for semantic answer similarity, trained on a new dataset consisting of name pairs of US-American public figures. As far as we are concerned, we provide the first dataset of co-referent name string pairs along with their similarities, which can be used for training.

Statistical inference with implicit SGD: proximal Robbins-Monro vs. Polyak-Ruppert

The implicit stochastic gradient descent (ISGD), a proximal version of SGD, is gaining interest in the literature due to its stability over (explicit) SGD. In this paper, we conduct an in-depth analysis of the two modes of ISGD for smooth convex functions, namely proximal Robbins-Monro (proxRM) and proximal Poylak-Ruppert (proxPR) procedures, for their use in statistical inference on model parameters. Specifically, we derive non-asymptotic point estimation error bounds of both proxRM and proxPR iterates and their limiting distributions, and propose on-line estimators of their asymptotic covariance matrices that require only a single run of ISGD. The latter estimators are used to construct valid confidence intervals for the model parameters. Our analysis is free of the generalized linear model assumption that has limited the preceding analyses, and employs feasible procedures. Our on-line covariance matrix estimators appear to be the first of this kind in the ISGD literature.

BackdoorBench: A Comprehensive Benchmark of Backdoor Learning

Backdoor learning is an emerging and important topic of studying the vulnerability of deep neural networks (DNNs). Many pioneering backdoor attack and defense methods are being proposed successively or concurrently, in the status of a rapid arms race. However, we find that the evaluations of new methods are often unthorough to verify their claims and real performance, mainly due to the rapid development, diverse settings, as well as the difficulties of implementation and reproducibility. Without thorough evaluations and comparisons, it is difficult to track the current progress and design the future development roadmap of the literature. To alleviate this dilemma, we build a comprehensive benchmark of backdoor learning, called BackdoorBench. It consists of an extensible modular based codebase (currently including implementations of 8 state-of-the-art (SOTA) attack and 9 SOTA defense algorithms), as well as a standardized protocol of a complete backdoor learning. We also provide comprehensive evaluations of every pair of 8 attacks against 9 defenses, with 5 poisoning ratios, based on 5 models and 4 datasets, thus 8,000 pairs of evaluations in total. We further present analysis from different perspectives about these 8,000 evaluations, studying the effects of attack against defense algorithms, poisoning ratio, model and dataset in backdoor learning. All codes and evaluations of BackdoorBench are publicly available at \url{https://backdoorbench.github.io}.

Review on Social Behavior Analysis of Laboratory Animals: From Methodologies to Applications

As the bridge between genetic and physiological aspects, animal behaviour analysis is one of the most significant topics in biology and ecological research. However, identifying, tracking and recording animal behaviour are labour intensive works that require professional knowledge. To mitigate the spend for annotating data, researchers turn to computer vision techniques for automatic label algorithms, since most of the data are recorded visually. In this work, we explore a variety of behaviour detection algorithms, covering traditional vision methods, statistical methods and deep learning methods. The objective of this work is to provide a thorough investigation of related work, furnishing biologists with a scratch of efficient animal behaviour detection methods. Apart from that, we also discuss the strengths and weaknesses of those algorithms to provide some insights for those who already delve into this field.

Machine Learning-based Biological Ageing Estimation Technologies: A Survey

In recent years, there are various methods of estimating Biological Age (BA) have been developed. Especially with the development of machine learning (ML), there are more and more types of BA predictions, and the accuracy has been greatly improved. The models for the estimation of BA play an important role in monitoring healthy aging, and could provide new tools to detect health status in the general population and give warnings to sub-healthy people. We will mainly review three age prediction methods by using ML. They are based on blood biomarkers, facial images, and structural neuroimaging features. For now, the model using blood biomarkers is the simplest, most direct, and most accurate method. The face image method is affected by various aspects such as race, environment, etc., the prediction accuracy is not very good, which cannot make a great contribution to the medical field. In summary, we are here to track the way forward in the era of big data for us and other potential general populations and show ways to leverage the vast amounts of data available today.

Tensor Recovery Based on A Novel Non-convex Function Minimax Logarithmic Concave Penalty Function

Non-convex relaxation methods have been widely used in tensor recovery problems, and compared with convex relaxation methods, can achieve better recovery results. In this paper, a new non-convex function, Minimax Logarithmic Concave Penalty (MLCP) function, is proposed, and some of its intrinsic properties are analyzed, among which it is interesting to find that the Logarithmic function is an upper bound of the MLCP function. The proposed function is generalized to tensor cases, yielding tensor MLCP and weighted tensor $L\gamma$-norm. Consider that its explicit solution cannot be obtained when applying it directly to the tensor recovery problem. Therefore, the corresponding equivalence theorems to solve such problem are given, namely, tensor equivalent MLCP theorem and equivalent weighted tensor $L\gamma$-norm theorem. In addition, we propose two EMLCP-based models for classic tensor recovery problems, namely low-rank tensor completion (LRTC) and tensor robust principal component analysis (TRPCA), and design proximal alternate linearization minimization (PALM) algorithms to solve them individually. Furthermore, based on the Kurdyka-{\L}ojasiwicz property, it is proved that the solution sequence of the proposed algorithm has finite length and converges to the critical point globally. Finally, Extensive experiments show that proposed algorithm achieve good results, and it is confirmed that the MLCP function is indeed better than the Logarithmic function in the minimization problem, which is consistent with the analysis of theoretical properties.

Multi-Variate Time Series Forecasting on Variable Subsets

We formulate a new inference task in the domain of multivariate time series forecasting (MTSF), called Variable Subset Forecast (VSF), where only a small subset of the variables is available during inference. Variables are absent during inference because of long-term data loss (eg. sensor failures) or high -> low-resource domain shift between train / test. To the best of our knowledge, robustness of MTSF models in presence of such failures, has not been studied in the literature. Through extensive evaluation, we first show that the performance of state of the art methods degrade significantly in the VSF setting. We propose a non-parametric, wrapper technique that can be applied on top any existing forecast models. Through systematic experiments across 4 datasets and 5 forecast models, we show that our technique is able to recover close to 95\% performance of the models even when only 15\% of the original variables are present.

Asymptotic-Preserving Neural Networks for multiscale hyperbolic models of epidemic spread

When investigating epidemic dynamics through differential models, the parameters needed to understand the phenomenon and to simulate forecast scenarios require a delicate calibration phase, often made even more challenging by the scarcity and uncertainty of the observed data reported by official sources. In this context, Physics-Informed Neural Networks (PINNs), by embedding the knowledge of the differential model that governs the physical phenomenon in the learning process, can effectively address the inverse and forward problem of data-driven learning and solving the corresponding epidemic problem. In many circumstances, however, the spatial propagation of an infectious disease is characterized by movements of individuals at different scales governed by multiscale PDEs. This reflects the heterogeneity of a region or territory in relation to the dynamics within cities and in neighboring zones. In presence of multiple scales, a direct application of PINNs generally leads to poor results due to the multiscale nature of the differential model in the loss function of the neural network. To allow the neural network to operate uniformly with respect to the small scales, it is desirable that the neural network satisfies an Asymptotic-Preservation (AP) property in the learning process. To this end, we consider a new class of AP Neural Networks (APNNs) for multiscale hyperbolic transport models of epidemic spread that, thanks to an appropriate AP formulation of the loss function, is capable to work uniformly at the different scales of the system. A series of numerical tests for different epidemic scenarios confirms the validity of the proposed approach, highlighting the importance of the AP property in the neural network when dealing with multiscale problems especially in presence of sparse and partially observed systems.

Language Models as Knowledge Embeddings

Knowledge embeddings (KE) represent a knowledge graph (KG) by embedding entities and relations into continuous vector spaces. Existing methods are mainly structure-based or description-based. Structure-based methods learn representations that preserve the inherent structure of KGs. They cannot well represent abundant long-tail entities in real-world KGs with limited structural information. Description-based methods leverage textual information and language models. Prior approaches in this direction barely outperform structure-based ones, and suffer from problems like expensive negative sampling and restrictive description demand. In this paper, we propose LMKE, which adopts Language Models to derive Knowledge Embeddings, aiming at both enriching representations of long-tail entities and solving problems of prior description-based methods. We formulate description-based KE learning with a contrastive learning framework to improve efficiency in training and evaluation. Experimental results show that LMKE achieves state-of-the-art performance on KE benchmarks of link prediction and triple classification, especially for long-tail entities.

Adversarial Self-Attention for Language Understanding

An ultimate language system aims at the high generalization and robustness when adapting to diverse scenarios. Unfortunately, the recent white hope pre-trained language models (PrLMs) barely escape from stacking excessive parameters to the over-parameterized Transformer architecture to achieve higher performances. This paper thus proposes \textit{Adversarial Self-Attention} mechanism (ASA), which adversarially reconstructs the Transformer attentions and facilitates model training from contaminated model structures, coupled with a fast and simple implementation for better PrLM building. We conduct comprehensive evaluation across a wide range of tasks on both pre-training and fine-tuning stages. For pre-training, ASA unfolds remarkable performance gain compared to regular training for longer periods. For fine-tuning, ASA-empowered models consistently outweigh naive models by a large margin considering both generalization and robustness.

Mitigating sampling bias in risk-based active learning via an EM algorithm

Risk-based active learning is an approach to developing statistical classifiers for online decision-support. In this approach, data-label querying is guided according to the expected value of perfect information for incipient data points. For SHM applications, the value of information is evaluated with respect to a maintenance decision process, and the data-label querying corresponds to the inspection of a structure to determine its health state. Sampling bias is a known issue within active-learning paradigms; this occurs when an active learning process over- or undersamples specific regions of a feature-space, thereby resulting in a training set that is not representative of the underlying distribution. This bias ultimately degrades decision-making performance, and as a consequence, results in unnecessary costs incurred. The current paper outlines a risk-based approach to active learning that utilises a semi-supervised Gaussian mixture model. The semi-supervised approach counteracts sampling bias by incorporating pseudo-labels for unlabelled data via an EM algorithm. The approach is demonstrated on a numerical example representative of the decision processes found in SHM.

Making Look-Ahead Active Learning Strategies Feasible with Neural Tangent Kernels

We propose a new method for approximating active learning acquisition strategies that are based on retraining with hypothetically-labeled candidate data points. Although this is usually infeasible with deep networks, we use the neural tangent kernel to approximate the result of retraining, and prove that this approximation works asymptotically even in an active learning setup -- approximating "look-ahead" selection criteria with far less computation required. This also enables us to conduct sequential active learning, i.e. updating the model in a streaming regime, without needing to retrain the model with SGD after adding each new data point. Moreover, our querying strategy, which better understands how the model's predictions will change by adding new data points in comparison to the standard ("myopic") criteria, beats other look-ahead strategies by large margins, and achieves equal or better performance compared to state-of-the-art methods on several benchmark datasets in pool-based active learning.

Functional Optimization Reinforcement Learning for Real-Time Bidding

Real-time bidding is the new paradigm of programmatic advertising. An advertiser wants to make the intelligent choice of utilizing a \textbf{Demand-Side Platform} to improve the performance of their ad campaigns. Existing approaches are struggling to provide a satisfactory solution for bidding optimization due to stochastic bidding behavior. In this paper, we proposed a multi-agent reinforcement learning architecture for RTB with functional optimization. We designed four agents bidding environment: three Lagrange-multiplier based functional optimization agents and one baseline agent (without any attribute of functional optimization) First, numerous attributes have been assigned to each agent, including biased or unbiased win probability, Lagrange multiplier, and click-through rate. In order to evaluate the proposed RTB strategy's performance, we demonstrate the results on ten sequential simulated auction campaigns. The results show that agents with functional actions and rewards had the most significant average winning rate and winning surplus, given biased and unbiased winning information respectively. The experimental evaluations show that our approach significantly improve the campaign's efficacy and profitability.

Generating Diverse Vocal Bursts with StyleGAN2 and MEL-Spectrograms

We describe our approach for the generative emotional vocal burst task (ExVo Generate) of the ICML Expressive Vocalizations Competition. We train a conditional StyleGAN2 architecture on mel-spectrograms of preprocessed versions of the audio samples. The mel-spectrograms generated by the model are then inverted back to the audio domain. As a result, our generated samples substantially improve upon the baseline provided by the competition from a qualitative and quantitative perspective for all emotions. More precisely, even for our worst-performing emotion (awe), we obtain an FAD of 1.76 compared to the baseline of 4.81 (as a reference, the FAD between the train/validation sets for awe is 0.776).

PLATON: Pruning Large Transformer Models with Upper Confidence Bound of Weight Importance

Large Transformer-based models have exhibited superior performance in various natural language processing and computer vision tasks. However, these models contain enormous amounts of parameters, which restrict their deployment to real-world applications. To reduce the model size, researchers prune these models based on the weights' importance scores. However, such scores are usually estimated on mini-batches during training, which incurs large variability/uncertainty due to mini-batch sampling and complicated training dynamics. As a result, some crucial weights could be pruned by commonly used pruning methods because of such uncertainty, which makes training unstable and hurts generalization. To resolve this issue, we propose PLATON, which captures the uncertainty of importance scores by upper confidence bound (UCB) of importance estimation. In particular, for the weights with low importance scores but high uncertainty, PLATON tends to retain them and explores their capacity. We conduct extensive experiments with several Transformer-based models on natural language understanding, question answering and image classification to validate the effectiveness of PLATON. Results demonstrate that PLATON manifests notable improvement under different sparsity levels. Our code is publicly available at https://github.com/QingruZhang/PLATON.

Integrating Machine Learning with Discrete Event Simulation for Improving Health Referral Processing in a Care Management Setting

Post-discharge care management coordinates patients' referrals to improve their health after being discharged from hospitals, especially elderly and chronically ill patients. In a care management setting, health referrals are processed by a specialized unit in the managed care organization (MCO), which interacts with many other entities including inpatient hospitals, insurance companies, and post-discharge care providers. In this paper, a machine-learning-guided discrete event simulation framework to improve health referrals processing is proposed. Random-forest-based prediction models are developed to predict the LOS and referral type. Two simulation models are constructed to represent the as-is configuration of the referral processing system and the intelligent system after incorporating the prediction functionality, respectively. By incorporating a prediction module for the referral processing system to plan and prioritize referrals, the overall performance was enhanced in terms of reducing the average referral creation delay time. This research will emphasize the role of post-discharge care management in improving health quality and reducing associated costs. Also, the paper demonstrates how to use integrated systems engineering methods for process improvement of complex healthcare systems.

Geometry Contrastive Learning on Heterogeneous Graphs

Self-supervised learning (especially contrastive learning) methods on heterogeneous graphs can effectively get rid of the dependence on supervisory data. Meanwhile, most existing representation learning methods embed the heterogeneous graphs into a single geometric space, either Euclidean or hyperbolic. This kind of single geometric view is usually not enough to observe the complete picture of heterogeneous graphs due to their rich semantics and complex structures. Under these observations, this paper proposes a novel self-supervised learning method, termed as Geometry Contrastive Learning (GCL), to better represent the heterogeneous graphs when supervisory data is unavailable. GCL views a heterogeneous graph from Euclidean and hyperbolic perspective simultaneously, aiming to make a strong merger of the ability of modeling rich semantics and complex structures, which is expected to bring in more benefits for downstream tasks. GCL maximizes the mutual information between two geometric views by contrasting representations at both local-local and local-global semantic levels. Extensive experiments on four benchmarks data sets show that the proposed approach outperforms the strong baselines, including both unsupervised methods and supervised methods, on three tasks, including node classification, node clustering and similarity search.

A Fast, Well-Founded Approximation to the Empirical Neural Tangent Kernel

Empirical neural tangent kernels (eNTKs) can provide a good understanding of a given network's representation: they are often far less expensive to compute and applicable more broadly than infinite width NTKs. For networks with O output units (e.g. an O-class classifier), however, the eNTK on N inputs is of size $NO \times NO$, taking $O((NO)^2)$ memory and up to $O((NO)^3)$ computation. Most existing applications have therefore used one of a handful of approximations yielding $N \times N$ kernel matrices, saving orders of magnitude of computation, but with limited to no justification. We prove that one such approximation, which we call "sum of logits", converges to the true eNTK at initialization for any network with a wide final "readout" layer. Our experiments demonstrate the quality of this approximation for various uses across a range of settings.

Value-Consistent Representation Learning for Data-Efficient Reinforcement Learning

Deep reinforcement learning (RL) algorithms suffer severe performance degradation when the interaction data is scarce, which limits their real-world application. Recently, visual representation learning has been shown to be effective and promising for boosting sample efficiency in RL. These methods usually rely on contrastive learning and data augmentation to train a transition model for state prediction, which is different from how the model is used in RL--performing value-based planning. Accordingly, the learned model may not be able to align well with the environment and generate consistent value predictions, especially when the state transition is not deterministic. To address this issue, we propose a novel method, called value-consistent representation learning (VCR), to learn representations that are directly related to decision-making. More specifically, VCR trains a model to predict the future state (also referred to as the ''imagined state'') based on the current one and a sequence of actions. Instead of aligning this imagined state with a real state returned by the environment, VCR applies a $Q$-value head on both states and obtains two distributions of action values. Then a distance is computed and minimized to force the imagined state to produce a similar action value prediction as that by the real state. We develop two implementations of the above idea for the discrete and continuous action spaces respectively. We conduct experiments on Atari 100K and DeepMind Control Suite benchmarks to validate their effectiveness for improving sample efficiency. It has been demonstrated that our methods achieve new state-of-the-art performance for search-free RL algorithms.

Visual Auditor: Interactive Visualization for Detection and Summarization of Model Biases

As machine learning (ML) systems become increasingly widespread, it is necessary to audit these systems for biases prior to their deployment. Recent research has developed algorithms for effectively identifying intersectional bias in the form of interpretable, underperforming subsets (or slices) of the data. However, these solutions and their insights are limited without a tool for visually understanding and interacting with the results of these algorithms. We propose Visual Auditor, an interactive visualization tool for auditing and summarizing model biases. Visual Auditor assists model validation by providing an interpretable overview of intersectional bias (bias that is present when examining populations defined by multiple features), details about relationships between problematic data slices, and a comparison between underperforming and overperforming data slices in a model. Our open-source tool runs directly in both computational notebooks and web browsers, making model auditing accessible and easily integrated into current ML development workflows. An observational user study in collaboration with domain experts at Fiddler AI highlights that our tool can help ML practitioners identify and understand model biases.

Learning the Ranking of Causal Effects with Confounded Data

Decision makers often want to identify the individuals for whom some intervention or treatment will be most effective in order to decide who to treat. In such cases, decision makers would ideally like to rank potential recipients of the treatment according to their individual causal effects. However, the historical data available to estimate the causal effects could be confounded, and as a result, accurately estimating the effects could be impossible. We propose a new and less restrictive assumption about historical data, called the ranking preservation assumption (RPA), under which the ranking of the individual effects can be consistently estimated even if the effects themselves cannot be accurately estimated. Importantly, we find that confounding can be helpful for the estimation of the causal-effect ranking when the confounding bias is larger for individuals with larger causal effects, and that even when this is not the case, any detrimental impact of confounding can be corrected with larger training data when the RPA is met. We then analytically show that the RPA can be met in a variety of scenarios, including common business applications such as online advertising and customer retention. We support this finding with an empirical example in the context of online advertising. The example also shows how to evaluate the decision making of a confounded model in practice. The main takeaway is that what might traditionally be considered "good" data for causal estimation (i.e., unconfounded data) may not be necessary to make good causal decisions, so treatment assignment methods may work better than we give them credit for in the presence of confounding.

Infinite Impulse Response Graph Neural Networks for Cyberattack Localization in Smart Grids

This study employs Infinite Impulse Response (IIR) Graph Neural Networks (GNN) to efficiently model the inherent graph network structure of the smart grid data to address the cyberattack localization problem. First, we numerically analyze the empirical frequency response of the Finite Impulse Response (FIR) and IIR graph filters (GFs) to approximate an ideal spectral response. We show that, for the same filter order, IIR GFs provide a better approximation to the desired spectral response and they also present the same level of approximation to a lower order GF due to their rational type filter response. Second, we propose an IIR GNN model to efficiently predict the presence of cyberattacks at the bus level. Finally, we evaluate the model under various cyberattacks at both sample-wise (SW) and bus-wise (BW) level, and compare the results with the existing architectures. It is experimentally verified that the proposed model outperforms the state-of-the-art FIR GNN model by 9.2% and 14% in terms of SW and BW localization, respectively.

Learning to learn online with neuromodulated synaptic plasticity in spiking neural networks

We propose that in order to harness our understanding of neuroscience toward machine learning, we must first have powerful tools for training brain-like models of learning. Although substantial progress has been made toward understanding the dynamics of learning in the brain, neuroscience-derived models of learning have yet to demonstrate the same performance capabilities as methods in deep learning such as gradient descent. Inspired by the successes of machine learning using gradient descent, we demonstrate that models of neuromodulated synaptic plasticity from neuroscience can be trained in Spiking Neural Networks (SNNs) with a framework of learning to learn through gradient descent to address challenging online learning problems. This framework opens a new path toward developing neuroscience inspired online learning algorithms.

Domain Generalization with Relaxed Instance Frequency-wise Normalization for Multi-device Acoustic Scene Classification

While using two-dimensional convolutional neural networks (2D-CNNs) in image processing, it is possible to manipulate domain information using channel statistics, and instance normalization has been a promising way to get domain-invariant features. However, unlike image processing, we analyze that domain-relevant information in an audio feature is dominant in frequency statistics rather than channel statistics. Motivated by our analysis, we introduce Relaxed Instance Frequency-wise Normalization (RFN): a plug-and-play, explicit normalization module along the frequency axis which can eliminate instance-specific domain discrepancy in an audio feature while relaxing undesirable loss of useful discriminative information. Empirically, simply adding RFN to networks shows clear margins compared to previous domain generalization approaches on acoustic scene classification and yields improved robustness for multiple audio devices. Especially, the proposed RFN won the DCASE2021 challenge TASK1A, low-complexity acoustic scene classification with multiple devices, with a clear margin, and RFN is an extended work of our technical report.

FetReg2021: A Challenge on Placental Vessel Segmentation and Registration in Fetoscopy

Fetoscopy laser photocoagulation is a widely adopted procedure for treating Twin-to-Twin Transfusion Syndrome (TTTS). The procedure involves photocoagulation pathological anastomoses to regulate blood exchange among twins. The procedure is particularly challenging due to the limited field of view, poor manoeuvrability of the fetoscope, poor visibility, and variability in illumination. These challenges may lead to increased surgery time and incomplete ablation. Computer-assisted intervention (CAI) can provide surgeons with decision support and context awareness by identifying key structures in the scene and expanding the fetoscopic field of view through video mosaicking. Research in this domain has been hampered by the lack of high-quality data to design, develop and test CAI algorithms. Through the Fetoscopic Placental Vessel Segmentation and Registration (FetReg2021) challenge, which was organized as part of the MICCAI2021 Endoscopic Vision challenge, we released the first largescale multicentre TTTS dataset for the development of generalized and robust semantic segmentation and video mosaicking algorithms. For this challenge, we released a dataset of 2060 images, pixel-annotated for vessels, tool, fetus and background classes, from 18 in-vivo TTTS fetoscopy procedures and 18 short video clips. Seven teams participated in this challenge and their model performance was assessed on an unseen test dataset of 658 pixel-annotated images from 6 fetoscopic procedures and 6 short clips. The challenge provided an opportunity for creating generalized solutions for fetoscopic scene understanding and mosaicking. In this paper, we present the findings of the FetReg2021 challenge alongside reporting a detailed literature review for CAI in TTTS fetoscopy. Through this challenge, its analysis and the release of multi-centre fetoscopic data, we provide a benchmark for future research in this field.

Black Box Optimization Using QUBO and the Cross Entropy Method

Black box optimization (BBO) can be used to optimize functions whose analytic form is unknown. A common approach to realize BBO is to learn a surrogate model which approximates the target black box function which can then be solved via white box optimization methods. In this paper we present our approach BOX-QUBO, where the surrogate model is a QUBO matrix. However, unlike in previous state-of-the-art approaches, this matrix is not trained entirely by regression, but mostly by classification between 'good' and 'bad' solutions. This better accounts for the low capacity of the QUBO matrix, resulting in significantly better solutions overall. We tested our approach against the state-of-the-art on four domains and in all of them BOX-QUBO showed significantly better results. A second contribution of this paper is the idea to also solve white box problems, i.e. problems which could be directly formulated as QUBO, by means of black box optimization in order to reduce the size of the QUBOs to their information-theoretic minimum. The experiments show that this significantly improves the results for MAX-$k$-SAT.

Multitask vocal burst modeling with ResNets and pre-trained paralinguistic Conformers

This technical report presents the modeling approaches used in our submission to the ICML Expressive Vocalizations Workshop & Competition multitask track (ExVo-MultiTask). We first applied image classification models of various sizes on mel-spectrogram representations of the vocal bursts, as is standard in sound event detection literature. Results from these models show an increase of 21.24% over the baseline system with respect to the harmonic mean of the task metrics, and comprise our team's main submission to the MultiTask track. We then sought to characterize the headroom in the MultiTask track by applying a large pre-trained Conformer model that previously achieved state-of-the-art results on paralinguistic tasks like speech emotion recognition and mask detection. We additionally investigated the relationship between the sub-tasks of emotional expression, country of origin, and age prediction, and discovered that the best performing models are trained as single-task models, questioning whether the problem truly benefits from a multitask setting.

RAPid-Learn: A Framework for Learning to Recover for Handling Novelties in Open-World Environments

We propose RAPid-Learn: Learning to Recover and Plan Again, a hybrid planning and learning method, to tackle the problem of adapting to sudden and unexpected changes in an agent's environment (i.e., novelties). RAPid-Learn is designed to formulate and solve modifications to a task's Markov Decision Process (MDPs) on-the-fly and is capable of exploiting domain knowledge to learn any new dynamics caused by the environmental changes. It is capable of exploiting the domain knowledge to learn action executors which can be further used to resolve execution impasses, leading to a successful plan execution. This novelty information is reflected in its updated domain model. We demonstrate its efficacy by introducing a wide variety of novelties in a gridworld environment inspired by Minecraft, and compare our algorithm with transfer learning baselines from the literature. Our method is (1) effective even in the presence of multiple novelties, (2) more sample efficient than transfer learning RL baselines, and (3) robust to incomplete model information, as opposed to pure symbolic planning approaches.

Arithmetic Circuits, Structured Matrices and (not so) Deep Learning

This survey presents a necessarily incomplete (and biased) overview of results at the intersection of arithmetic circuit complexity, structured matrices and deep learning. Recently there has been some research activity in replacing unstructured weight matrices in neural networks by structured ones (with the aim of reducing the size of the corresponding deep learning models). Most of this work has been experimental and in this survey, we formalize the research question and show how a recent work that combines arithmetic circuit complexity, structured matrices and deep learning essentially answers this question. This survey is targeted at complexity theorists who might enjoy reading about how tools developed in arithmetic circuit complexity helped design (to the best of our knowledge) a new family of structured matrices, which in turn seem well-suited for applications in deep learning. However, we hope that folks primarily interested in deep learning would also appreciate the connections to complexity theory.

Variational Bayesian inference for CP tensor completion with side information

We propose a message passing algorithm, based on variational Bayesian inference, for low-rank tensor completion with automatic rank determination in the canonical polyadic format when additional side information (SI) is given. The SI comes in the form of low-dimensional subspaces the contain the fiber spans of the tensor (columns, rows, tubes, etc.). We validate the regularization properties induced by SI with extensive numerical experiments on synthetic and real-world data and present the results about tensor recovery and rank determination. The results show that the number of samples required for successful completion is significantly reduced in the presence of SI. We also discuss the origin of a bump in the phase transition curves that exists when the dimensionality of SI is comparable with that of the tensor.

A Novel Approach For Analysis of Distributed Acoustic Sensing System Based on Deep Transfer Learning

Distributed acoustic sensors (DAS) are effective apparatus which are widely used in many application areas for recording signals of various events with very high spatial resolution along the optical fiber. To detect and recognize the recorded events properly, advanced signal processing algorithms with high computational demands are crucial. Convolutional neural networks are highly capable tools for extracting spatial information and very suitable for event recognition applications in DAS. Long-short term memory (LSTM) is an effective instrument for processing sequential data. In this study, we proposed a multi-input multi-output, two stage feature extraction methodology that combines the capabilities of these neural network architectures with transfer learning to classify vibrations applied to an optical fiber by a piezo transducer. First, we extracted the differential amplitude and phase information from the Phase-OTDR recordings and stored them in a temporal-spatial data matrix. Then, we used a state-of-the-art pre-trained CNN without dense layers as a feature extractor in the first stage. In the second stage, we used LSTMs to further analyze the features extracted by the CNN. Finally, we used a dense layer to classify the extracted features. To observe the effect of the utilized CNN architecture, we tested our model with five state-of-the art pre-trained models (VGG-16, ResNet-50, DenseNet-121, MobileNet and Inception-v3). The results show that using the VGG-16 architecture in our framework manages to obtain 100% classification accuracy in 50 trainings and got the best results on our Phase-OTDR dataset. Outcomes of this study indicate that the pre-trained CNNs combined with LSTM are very suitable for the analysis of differential amplitude and phase information, represented in a temporal spatial data matrix which is promising for event recognition operations in DAS applications.

Analyzing the Effects of Classifier Lipschitzness on Explainers

Machine learning methods are getting increasingly better at making predictions, but at the same time they are also becoming more complicated and less transparent. As a result, explainers are often relied on to provide interpretability to these black-box prediction models. As crucial diagnostics tools, it is important that these explainers themselves are reliable. In this paper we focus on one particular aspect of reliability, namely that an explainer should give similar explanations for similar data inputs. We formalize this notion by introducing and defining explainer astuteness, analogous to astuteness of classifiers. Our formalism is inspired by the concept of probabilistic Lipschitzness, which captures the probability of local smoothness of a function. For a variety of explainers (e.g., SHAP, RISE, CXPlain), we provide lower bound guarantees on the astuteness of these explainers given the Lipschitzness of the prediction function. These theoretical results imply that locally smooth prediction functions lend themselves to locally robust explanations. We evaluate these results empirically on simulated as well as real datasets.

Risk-averse Contextual Multi-armed Bandit Problem with Linear Payoffs

In this paper we consider the contextual multi-armed bandit problem for linear payoffs under a risk-averse criterion. At each round, contexts are revealed for each arm, and the decision maker chooses one arm to pull and receives the corresponding reward. In particular, we consider mean-variance as the risk criterion, and the best arm is the one with the largest mean-variance reward. We apply the Thompson Sampling algorithm for the disjoint model, and provide a comprehensive regret analysis for a variant of the proposed algorithm. For $T$ rounds, $K$ actions, and $d$-dimensional feature vectors, we prove a regret bound of $O((1+\rho+\frac{1}{\rho}) d\ln T \ln \frac{K}{\delta}\sqrt{d K T^{1+2\epsilon} \ln \frac{K}{\delta} \frac{1}{\epsilon}})$ that holds with probability $1-\delta$ under the mean-variance criterion with risk tolerance $\rho$, for any $0<\epsilon<\frac{1}{2}$, $0<\delta<1$. The empirical performance of our proposed algorithms is demonstrated via a portfolio selection problem.

Vibration fault detection in wind turbines based on normal behaviour models without feature engineering

Most wind turbines are remotely monitored 24/7 to allow for an early detection of operation problems and developing damage. We present a new fault detection method for vibration-monitored drivetrains that does not require any feature engineering. Our method relies on a simple model architecture to enable a straightforward implementation in practice. We propose to apply convolutional autoencoders for identifying and extracting the most relevant features from the half spectrum in an automated manner, saving time and effort. Thereby, a spectral model of the normal vibration response is learnt for the monitored component from past measurements. We demonstrate that the model can successfully distinguish damaged from healthy components and detect a damaged generator bearing and damaged gearbox parts from their vibration responses. Using measurements from commercial wind turbines and a test rig, we show that vibration-based fault detection in wind turbine drivetrains can be performed without the usual upfront definition of spectral features. Another advantage of the presented method is that the entire half spectrum is monitored instead of the usual focus on monitoring individual frequencies and harmonics.

Value Function Decomposition for Iterative Design of Reinforcement Learning Agents

Designing reinforcement learning (RL) agents is typically a difficult process that requires numerous design iterations. Learning can fail for a multitude of reasons, and standard RL methods provide too few tools to provide insight into the exact cause. In this paper, we show how to integrate value decomposition into a broad class of actor-critic algorithms and use it to assist in the iterative agent-design process. Value decomposition separates a reward function into distinct components and learns value estimates for each. These value estimates provide insight into an agent's learning and decision-making process and enable new training methods to mitigate common problems. As a demonstration, we introduce SAC-D, a variant of soft actor-critic (SAC) adapted for value decomposition. SAC-D maintains similar performance to SAC, while learning a larger set of value predictions. We also introduce decomposition-based tools that exploit this information, including a new reward influence metric, which measures each reward component's effect on agent decision-making. Using these tools, we provide several demonstrations of decomposition's use in identifying and addressing problems in the design of both environments and agents. Value decomposition is broadly applicable and easy to incorporate into existing algorithms and workflows, making it a powerful tool in an RL practitioner's toolbox.

Gated Domain Units for Multi-source Domain Generalization

Distribution shift (DS) is a common problem that deteriorates the performance of learning machines. To overcome this problem, we postulate that real-world distributions are composed of elementary distributions that remain invariant across different domains. We call this an invariant elementary distribution (I.E.D.) assumption. This invariance thus enables knowledge transfer to unseen domains. To exploit this assumption in domain generalization (DG), we developed a modular neural network layer that consists of Gated Domain Units (GDUs). Each GDU learns an embedding of an individual elementary domain that allows us to encode the domain similarities during the training. During inference, the GDUs compute similarities between an observation and each of the corresponding elementary distributions which are then used to form a weighted ensemble of learning machines. Because our layer is trained with backpropagation, it can be easily integrated into existing deep learning frameworks. Our evaluation on Digits5, ECG, Camelyon17, iWildCam, and FMoW shows a significant improvement in the performance on out-of-training target domains without any access to data from the target domains. This finding supports the validity of the I.E.D. assumption in real-world data distributions.

Joint Representation Training in Sequential Tasks with Shared Structure

Classical theory in reinforcement learning (RL) predominantly focuses on the single task setting, where an agent learns to solve a task through trial-and-error experience, given access to data only from that task. However, many recent empirical works have demonstrated the significant practical benefits of leveraging a joint representation trained across multiple, related tasks. In this work we theoretically analyze such a setting, formalizing the concept of task relatedness as a shared state-action representation that admits linear dynamics in all the tasks. We introduce the Shared-MatrixRL algorithm for the setting of Multitask MatrixRL. In the presence of $P$ episodic tasks of dimension $d$ sharing a joint $r \ll d$ low-dimensional representation, we show the regret on the the $P$ tasks can be improved from $O(PHd\sqrt{NH})$ to $O((Hd\sqrt{rP} + HP\sqrt{rd})\sqrt{NH})$ over $N$ episodes of horizon $H$. These gains coincide with those observed in other linear models in contextual bandits and RL. In contrast with previous work that have studied multi task RL in other function approximation models, we show that in the presence of bilinear optimization oracle and finite state action spaces there exists a computationally efficient algorithm for multitask MatrixRL via a reduction to quadratic programming. We also develop a simple technique to shave off a $\sqrt{H}$ factor from the regret upper bounds of some episodic linear problems.

ZSON: Zero-Shot Object-Goal Navigation using Multimodal Goal Embeddings

We present a scalable approach for learning open-world object-goal navigation (ObjectNav) -- the task of asking a virtual robot (agent) to find any instance of an object in an unexplored environment (e.g., "find a sink"). Our approach is entirely zero-shot -- i.e., it does not require ObjectNav rewards or demonstrations of any kind. Instead, we train on the image-goal navigation (ImageNav) task, in which agents find the location where a picture (i.e., goal image) was captured. Specifically, we encode goal images into a multimodal, semantic embedding space to enable training semantic-goal navigation (SemanticNav) agents at scale in unannotated 3D environments (e.g., HM3D). After training, SemanticNav agents can be instructed to find objects described in free-form natural language (e.g., "sink", "bathroom sink", etc.) by projecting language goals into the same multimodal, semantic embedding space. As a result, our approach enables open-world ObjectNav. We extensively evaluate our agents on three ObjectNav datasets (Gibson, HM3D, and MP3D) and observe absolute improvements in success of 4.2% - 20.0% over existing zero-shot methods. For reference, these gains are similar or better than the 5% improvement in success between the Habitat 2020 and 2021 ObjectNav challenge winners. In an open-world setting, we discover that our agents can generalize to compound instructions with a room explicitly mentioned (e.g., "Find a kitchen sink") and when the target room can be inferred (e.g., "Find a sink and a stove").

Predicting the Stability of Hierarchical Triple Systems with Convolutional Neural Networks

Understanding the long-term evolution of hierarchical triple systems is challenging due to its inherent chaotic nature, and it requires computationally expensive simulations. Here we propose a convolutional neural network model to predict the stability of hierarchical triples by looking at their evolution during the first $5 \times 10^5$ inner binary orbits. We employ the regularized few-body code \textsc{tsunami} to simulate $5\times 10^6$ hierarchical triples, from which we generate a large training and test dataset. We develop twelve different network configurations that use different combinations of the triples' orbital elements and compare their performances. Our best model uses 6 time-series, namely, the semimajor axes ratio, the inner and outer eccentricities, the mutual inclination and the arguments of pericenter. This model achieves an area under the curve of over $95\%$ and informs of the relevant parameters to study triple systems stability. All trained models are made publicly available, allowing to predict the stability of hierarchical triple systems $200$ times faster than pure $N$-body methods.

Debiasing Learning for Membership Inference Attacks Against Recommender Systems

Learned recommender systems may inadvertently leak information about their training data, leading to privacy violations. We investigate privacy threats faced by recommender systems through the lens of membership inference. In such attacks, an adversary aims to infer whether a user's data is used to train the target recommender. To achieve this, previous work has used a shadow recommender to derive training data for the attack model, and then predicts the membership by calculating difference vectors between users' historical interactions and recommended items. State-of-the-art methods face two challenging problems: (1) training data for the attack model is biased due to the gap between shadow and target recommenders, and (2) hidden states in recommenders are not observational, resulting in inaccurate estimations of difference vectors. To address the above limitations, we propose a Debiasing Learning for Membership Inference Attacks against recommender systems (DL-MIA) framework that has four main components: (1) a difference vector generator, (2) a disentangled encoder, (3) a weight estimator, and (4) an attack model. To mitigate the gap between recommenders, a variational auto-encoder (VAE) based disentangled encoder is devised to identify recommender invariant and specific features. To reduce the estimation bias, we design a weight estimator, assigning a truth-level score for each difference vector to indicate estimation accuracy. We evaluate DL-MIA against both general recommenders and sequential recommenders on three real-world datasets. Experimental results show that DL-MIA effectively alleviates training and estimation biases simultaneously, and achieves state-of-the-art attack performance.

Data Leakage in Federated Averaging

Recent attacks have shown that user data can be recovered from FedSGD updates, thus breaking privacy. However, these attacks are of limited practical relevance as federated learning typically uses the FedAvg algorithm. Compared to FedSGD, recovering data from FedAvg updates is much harder as: (i) the updates are computed at unobserved intermediate network weights, (ii) a large number of batches are used, and (iii) labels and network weights vary simultaneously across client steps. In this work, we propose a new optimization-based attack which successfully attacks FedAvg by addressing the above challenges. First, we solve the optimization problem using automatic differentiation that forces a simulation of the client's update that generates the unobserved parameters for the recovered labels and inputs to match the received client update. Second, we address the large number of batches by relating images from different epochs with a permutation invariant prior. Third, we recover the labels by estimating the parameters of existing FedSGD attacks at every FedAvg step. On the popular FEMNIST dataset, we demonstrate that on average we successfully recover >45% of the client's images from realistic FedAvg updates computed on 10 local epochs of 10 batches each with 5 images, compared to only <10% using the baseline. Our findings show many real-world federated learning implementations based on FedAvg are vulnerable.

On Certifying and Improving Generalization to Unseen Domains

Domain Generalization (DG) aims to learn models whose performance remains high on unseen domains encountered at test-time by using data from multiple related source domains. Many existing DG algorithms reduce the divergence between source distributions in a representation space to potentially align the unseen domain close to the sources. This is motivated by the analysis that explains generalization to unseen domains using distributional distance (such as the Wasserstein distance) to the sources. However, due to the openness of the DG objective, it is challenging to evaluate DG algorithms comprehensively using a few benchmark datasets. In particular, we demonstrate that the accuracy of the models trained with DG methods varies significantly across unseen domains, generated from popular benchmark datasets. This highlights that the performance of DG methods on a few benchmark datasets may not be representative of their performance on unseen domains in the wild. To overcome this roadblock, we propose a universal certification framework based on distributionally robust optimization (DRO) that can efficiently certify the worst-case performance of any DG method. This enables a data-independent evaluation of a DG method complementary to the empirical evaluations on benchmark datasets. Furthermore, we propose a training algorithm that can be used with any DG method to provably improve their certified performance. Our empirical evaluation demonstrates the effectiveness of our method at significantly improving the worst-case loss (i.e., reducing the risk of failure of these models in the wild) without incurring a significant performance drop on benchmark datasets.

From Tensor Network Quantum States to Tensorial Recurrent Neural Networks

We show that any matrix product state (MPS) can be exactly represented by a recurrent neural network (RNN) with a linear memory update. We generalize this RNN architecture to 2D lattices using a multilinear memory update. It supports perfect sampling and wave function evaluation in polynomial time, and can represent an area law of entanglement entropy. Numerical evidence shows that it can encode the wave function using a bond dimension lower by orders of magnitude when compared to MPS, with an accuracy that can be systematically improved by increasing the bond dimension.

Out of distribution robustness with pre-trained Bayesian neural networks

We develop ShiftMatch, a new training-data-dependent likelihood for out of distribution (OOD) robustness in Bayesian neural networks (BNNs). ShiftMatch is inspired by the training-data-dependent "EmpCov" priors from Izmailov et al. (2021a) and efficiently matches test-time spatial correlations to those at training time. Critically, ShiftMatch is designed to leave neural network training unchanged, allowing it to use publically available samples from pretrained BNNs. Using pre-trained HMC samples, ShiftMatch gives strong performance improvements on CIFAR-10-C, outperforms EmpCov priors, and is perhaps the first Bayesian method capable of convincingly outperforming plain deep ensembles. ShiftMatch can be integrated with non-Bayesian methods like deep ensembles, where it offers smaller, but still considerable, performance improvements. Overall, Bayesian ShiftMatch gave slightly better accuracy than ensembles with ShiftMatch, though they both had very similar log-likelihoods.

Quantifying Inherent Randomness in Machine Learning Algorithms

Most machine learning (ML) algorithms have several stochastic elements, and their performances are affected by these sources of randomness. This paper uses an empirical study to systematically examine the effects of two sources: randomness in model training and randomness in the partitioning of a dataset into training and test subsets. We quantify and compare the magnitude of the variation in predictive performance for the following ML algorithms: Random Forests (RFs), Gradient Boosting Machines (GBMs), and Feedforward Neural Networks (FFNNs). Among the different algorithms, randomness in model training causes larger variation for FFNNs compared to tree-based methods. This is to be expected as FFNNs have more stochastic elements that are part of their model initialization and training. We also found that random splitting of datasets leads to higher variation compared to the inherent randomness from model training. The variation from data splitting can be a major issue if the original dataset has considerable heterogeneity. Keywords: Model Training, Reproducibility, Variation

Megapixel Image Generation with Step-Unrolled Denoising Autoencoders

An ongoing trend in generative modelling research has been to push sample resolutions higher whilst simultaneously reducing computational requirements for training and sampling. We aim to push this trend further via the combination of techniques - each component representing the current pinnacle of efficiency in their respective areas. These include vector-quantized GAN (VQ-GAN), a vector-quantization (VQ) model capable of high levels of lossy - but perceptually insignificant - compression; hourglass transformers, a highly scaleable self-attention model; and step-unrolled denoising autoencoders (SUNDAE), a non-autoregressive (NAR) text generative model. Unexpectedly, our method highlights weaknesses in the original formulation of hourglass transformers when applied to multidimensional data. In light of this, we propose modifications to the resampling mechanism, applicable in any task applying hierarchical transformers to multidimensional data. Additionally, we demonstrate the scalability of SUNDAE to long sequence lengths - four times longer than prior work. Our proposed framework scales to high-resolutions ($1024 \times 1024$) and trains quickly (2-4 days). Crucially, the trained model produces diverse and realistic megapixel samples in approximately 2 seconds on a consumer-grade GPU (GTX 1080Ti). In general, the framework is flexible: supporting an arbitrary number of sampling steps, sample-wise self-stopping, self-correction capabilities, conditional generation, and a NAR formulation that allows for arbitrary inpainting masks. We obtain FID scores of 10.56 on FFHQ256 - close to the original VQ-GAN in less than half the sampling steps - and 21.85 on FFHQ1024 in only 100 sampling steps.

Segmentation-free PVC for Cardiac SPECT using a Densely-connected Multi-dimensional Dynamic Network

In nuclear imaging, limited resolution causes partial volume effects (PVEs) that affect image sharpness and quantitative accuracy. Partial volume correction (PVC) methods incorporating high-resolution anatomical information from CT or MRI have been demonstrated to be effective. However, such anatomical-guided methods typically require tedious image registration and segmentation steps. Accurately segmented organ templates are also hard to obtain, particularly in cardiac SPECT imaging, due to the lack of hybrid SPECT/CT scanners with high-end CT and associated motion artifacts. Slight mis-registration/mis-segmentation would result in severe degradation in image quality after PVC. In this work, we develop a deep-learning-based method for fast cardiac SPECT PVC without anatomical information and associated organ segmentation. The proposed network involves a densely-connected multi-dimensional dynamic mechanism, allowing the convolutional kernels to be adapted based on the input images, even after the network is fully trained. Intramyocardial blood volume (IMBV) is introduced as an additional clinical-relevant loss function for network optimization. The proposed network demonstrated promising performance on 28 canine studies acquired on a GE Discovery NM/CT 570c dedicated cardiac SPECT scanner with a 64-slice CT using Technetium-99m-labeled red blood cells. This work showed that the proposed network with densely-connected dynamic mechanism produced superior results compared with the same network without such mechanism. Results also showed that the proposed network without anatomical information could produce images with statistically comparable IMBV measurements to the images generated by anatomical-guided PVC methods, which could be helpful in clinical translation.

HANF: Hyperparameter And Neural Architecture Search in Federated Learning

Automated machine learning (AutoML) is an important step to make machine learning models being widely applied to solve real world problems. Despite numerous research advancement, machine learning methods are not fully utilized by industries mainly due to their data privacy and security regulations, high cost involved in storing and computing increasing amount of data at central location and most importantly lack of expertise. Hence, we introduce a novel framework, HANF - $\textbf{H}$yperparameter $\textbf{A}$nd $\textbf{N}$eural architecture search in $\textbf{F}$ederated learning as a step towards building an AutoML framework for data distributed across several data owner servers without any need for bringing the data to a central location. HANF jointly optimizes a neural architecture and non-architectural hyperparameters of a learning algorithm using gradient-based neural architecture search and $n$-armed bandit approach respectively in data distributed setting. We show that HANF efficiently finds the optimized neural architecture and also tunes the hyperparameters on data owner servers. Additionally, HANF can be applied in both, federated and non-federated settings. Empirically, we show that HANF converges towards well-suited architectures and non-architectural hyperparameter-sets using image-classification tasks.

Achievement and Fragility of Long-term Equitability

Equipping current decision-making tools with notions of fairness, equitability, or other ethically motivated outcomes, is one of the top priorities in recent research efforts in machine learning, AI, and optimization. In this paper, we investigate how to allocate limited resources to {locally interacting} communities in a way to maximize a pertinent notion of equitability. In particular, we look at the dynamic setting where the allocation is repeated across multiple periods (e.g., yearly), the local communities evolve in the meantime (driven by the provided allocation), and the allocations are modulated by feedback coming from the communities themselves. We employ recent mathematical tools stemming from data-driven feedback online optimization, by which communities can learn their (possibly unknown) evolution, satisfaction, as well as they can share information with the deciding bodies. We design dynamic policies that converge to an allocation that maximize equitability in the long term. We further demonstrate our model and methodology with realistic examples of healthcare and education subsidies design in Sub-Saharian countries. One of the key empirical takeaways from our setting is that long-term equitability is fragile, in the sense that it can be easily lost when deciding bodies weigh in other factors (e.g., equality in allocation) in the allocation strategy. Moreover, a naive compromise, while not providing significant advantage to the communities, can promote inequality in social outcomes.

Source Localization of Graph Diffusion via Variational Autoencoders for Graph Inverse Problems

Graph diffusion problems such as the propagation of rumors, computer viruses, or smart grid failures are ubiquitous and societal. Hence it is usually crucial to identify diffusion sources according to the current graph diffusion observations. Despite its tremendous necessity and significance in practice, source localization, as the inverse problem of graph diffusion, is extremely challenging as it is ill-posed: different sources may lead to the same graph diffusion patterns. Different from most traditional source localization methods, this paper focuses on a probabilistic manner to account for the uncertainty of different candidate sources. Such endeavors require overcoming challenges including 1) the uncertainty in graph diffusion source localization is hard to be quantified; 2) the complex patterns of the graph diffusion sources are difficult to be probabilistically characterized; 3) the generalization under any underlying diffusion patterns is hard to be imposed. To solve the above challenges, this paper presents a generic framework: Source Localization Variational AutoEncoder (SL-VAE) for locating the diffusion sources under arbitrary diffusion patterns. Particularly, we propose a probabilistic model that leverages the forward diffusion estimation model along with deep generative models to approximate the diffusion source distribution for quantifying the uncertainty. SL-VAE further utilizes prior knowledge of the source-observation pairs to characterize the complex patterns of diffusion sources by a learned generative prior. Lastly, a unified objective that integrates the forward diffusion estimation model is derived to enforce the model to generalize under arbitrary diffusion patterns. Extensive experiments are conducted on 7 real-world datasets to demonstrate the superiority of SL-VAE in reconstructing the diffusion sources by excelling other methods on average 20% in AUC score.

ModLaNets: Learning Generalisable Dynamics via Modularity and Physical Inductive Bias

Deep learning models are able to approximate one specific dynamical system but struggle at learning generalisable dynamics, where dynamical systems obey the same laws of physics but contain different numbers of elements (e.g., double- and triple-pendulum systems). To relieve this issue, we proposed the Modular Lagrangian Network (ModLaNet), a structural neural network framework with modularity and physical inductive bias. This framework models the energy of each element using modularity and then construct the target dynamical system via Lagrangian mechanics. Modularity is beneficial for reusing trained networks and reducing the scale of networks and datasets. As a result, our framework can learn from the dynamics of simpler systems and extend to more complex ones, which is not feasible using other relevant physics-informed neural networks. We examine our framework for modelling double-pendulum or three-body systems with small training datasets, where our models achieve the best data efficiency and accuracy performance compared with counterparts. We also reorganise our models as extensions to model multi-pendulum and multi-body systems, demonstrating the intriguing reusable feature of our framework.

On the Importance of Application-Grounded Experimental Design for Evaluating Explainable ML Methods

Machine Learning (ML) models now inform a wide range of human decisions, but using ``black box'' models carries risks such as relying on spurious correlations or errant data. To address this, researchers have proposed methods for supplementing models with explanations of their predictions. However, robust evaluations of these methods' usefulness in real-world contexts have remained elusive, with experiments tending to rely on simplified settings or proxy tasks. We present an experimental study extending a prior explainable ML evaluation experiment and bringing the setup closer to the deployment setting by relaxing its simplifying assumptions. Our empirical study draws dramatically different conclusions than the prior work, highlighting how seemingly trivial experimental design choices can yield misleading results. Beyond the present experiment, we believe this work holds lessons about the necessity of situating the evaluation of any ML method and choosing appropriate tasks, data, users, and metrics to match the intended deployment contexts.

How to train accurate BNNs for embedded systems?

A key enabler of deploying convolutional neural networks on resource-constrained embedded systems is the binary neural network (BNN). BNNs save on memory and simplify computation by binarizing both features and weights. Unfortunately, binarization is inevitably accompanied by a severe decrease in accuracy. To reduce the accuracy gap between binary and full-precision networks, many repair methods have been proposed in the recent past, which we have classified and put into a single overview in this chapter. The repair methods are divided into two main branches, training techniques and network topology changes, which can further be split into smaller categories. The latter category introduces additional cost (energy consumption or additional area) for an embedded system, while the former does not. From our overview, we observe that progress has been made in reducing the accuracy gap, but BNN papers are not aligned on what repair methods should be used to get highly accurate BNNs. Therefore, this chapter contains an empirical review that evaluates the benefits of many repair methods in isolation over the ResNet-20\&CIFAR10 and ResNet-18\&CIFAR100 benchmarks. We found three repair categories most beneficial: feature binarizer, feature normalization, and double residual. Based on this review we discuss future directions and research opportunities. We sketch the benefit and costs associated with BNNs on embedded systems because it remains to be seen whether BNNs will be able to close the accuracy gap while staying highly energy-efficient on resource-constrained embedded systems.

DeepAL for Regression Using $ε$-weighted Hybrid Query Strategy

Designing an inexpensive approximate surrogate model that captures the salient features of an expensive high-fidelity behavior is a prevalent approach in design optimization. In recent times, Deep Learning (DL) models are being used as a promising surrogate computational model for engineering problems. However, the main challenge in creating a DL-based surrogate is to simulate/label a large number of design points, which is time-consuming for computationally costly and/or high-dimensional engineering problems. In the present work, we propose a novel sampling technique by combining the active learning (AL) method with DL. We call this method $\epsilon$-weighted hybrid query strategy ($\epsilon$-HQS) , which focuses on the evaluation of the surrogate at each learning iteration and provides an estimate of the failure probability of the surrogate in the Design Space. By reusing already collected training and test data, the learned failure probability guides the next iteration's sampling process to the region of the high probability of failure. During the empirical evaluation, better accuracy of the surrogate was observed in comparison to other methods of sample selection. We empirically evaluated this method in two different engineering design domains, finite element based static stress analysis of submarine pressure vessel(computationally costly process) and second submarine propeller design( high dimensional problem). https://github.com/vardhah/epsilon_weighted_Hybrid_Query_Strategy

Learning sparse features can lead to overfitting in neural networks

It is widely believed that the success of deep networks lies in their ability to learn a meaningful representation of the features of the data. Yet, understanding when and how this feature learning improves performance remains a challenge: for example, it is beneficial for modern architectures trained to classify images, whereas it is detrimental for fully-connected networks trained for the same task on the same data. Here we propose an explanation for this puzzle, by showing that feature learning can perform worse than lazy training (via random feature kernel or the NTK) as the former can lead to a sparser neural representation. Although sparsity is known to be essential for learning anisotropic data, it is detrimental when the target function is constant or smooth along certain directions of input space. We illustrate this phenomenon in two settings: (i) regression of Gaussian random functions on the d-dimensional unit sphere and (ii) classification of benchmark datasets of images. For (i), we compute the scaling of the generalization error with number of training points, and show that methods that do not learn features generalize better, even when the dimension of the input space is large. For (ii), we show empirically that learning features can indeed lead to sparse and thereby less smooth representations of the image predictors. This fact is plausibly responsible for deteriorating the performance, which is known to be correlated with smoothness along diffeomorphisms.

Bugs in Machine Learning-based Systems: A Faultload Benchmark

The rapid escalation of applying Machine Learning (ML) in various domains has led to paying more attention to the quality of ML components. There is then a growth of techniques and tools aiming at improving the quality of ML components and integrating them into the ML-based system safely. Although most of these tools use bugs' lifecycle, there is no standard benchmark of bugs to assess their performance, compare them and discuss their advantages and weaknesses. In this study, we firstly investigate the reproducibility and verifiability of the bugs in ML-based systems and show the most important factors in each one. Then, we explore the challenges of generating a benchmark of bugs in ML-based software systems and provide a bug benchmark namely defect4ML that satisfies all criteria of standard benchmark, i.e. relevance, reproducibility, fairness, verifiability, and usability. This faultload benchmark contains 113 bugs reported by ML developers on GitHub and Stack Overflow, using two of the most popular ML frameworks: TensorFlow and Keras. defect4ML also addresses important challenges in Software Reliability Engineering of ML-based software systems, like: 1) fast changes in frameworks, by providing various bugs for different versions of frameworks, 2) code portability, by delivering similar bugs in different ML frameworks, 3) bug reproducibility, by providing fully reproducible bugs with complete information about required dependencies and data, and 4) lack of detailed information on bugs, by presenting links to the bugs' origins. defect4ML can be of interest to ML-based systems practitioners and researchers to assess their testing tools and techniques.

Analyzing the impact of SARS-CoV-2 variants on respiratory sound signals

The COVID-19 outbreak resulted in multiple waves of infections that have been associated with different SARS-CoV-2 variants. Studies have reported differential impact of the variants on respiratory health of patients. We explore whether acoustic signals, collected from COVID-19 subjects, show computationally distinguishable acoustic patterns suggesting a possibility to predict the underlying virus variant. We analyze the Coswara dataset which is collected from three subject pools, namely, i) healthy, ii) COVID-19 subjects recorded during the delta variant dominant period, and iii) data from COVID-19 subjects recorded during the omicron surge. Our findings suggest that multiple sound categories, such as cough, breathing, and speech, indicate significant acoustic feature differences when comparing COVID-19 subjects with omicron and delta variants. The classification areas-under-the-curve are significantly above chance for differentiating subjects infected by omicron from those infected by delta. Using a score fusion from multiple sound categories, we obtained an area-under-the-curve of 89% and 52.4% sensitivity at 95% specificity. Additionally, a hierarchical three class approach was used to classify the acoustic data into healthy and COVID-19 positive, and further COVID-19 subjects into delta and omicron variants providing high level of 3-class classification accuracy. These results suggest new ways for designing sound based COVID-19 diagnosis approaches.

PSP: Million-level Protein Sequence Dataset for Protein Structure Prediction

Proteins are essential component of human life and their structures are important for function and mechanism analysis. Recent work has shown the potential of AI-driven methods for protein structure prediction. However, the development of new models is restricted by the lack of dataset and benchmark training procedure. To the best of our knowledge, the existing open source datasets are far less to satisfy the needs of modern protein sequence-structure related research. To solve this problem, we present the first million-level protein structure prediction dataset with high coverage and diversity, named as PSP. This dataset consists of 570k true structure sequences (10TB) and 745k complementary distillation sequences (15TB). We provide in addition the benchmark training procedure for SOTA protein structure prediction model on this dataset. We validate the utility of this dataset for training by participating CAMEO contest in which our model won the first place. We hope our PSP dataset together with the training benchmark can enable a broader community of AI/biology researchers for AI-driven protein related research.

Iterative Sound Source Localization for Unknown Number of Sources

Sound source localization aims to seek the direction of arrival (DOA) of all sound sources from the observed multi-channel audio. For the practical problem of unknown number of sources, existing localization algorithms attempt to predict a likelihood-based coding (i.e., spatial spectrum) and employ a pre-determined threshold to detect the source number and corresponding DOA value. However, these threshold-based algorithms are not stable since they are limited by the careful choice of threshold. To address this problem, we propose an iterative sound source localization approach called ISSL, which can iteratively extract each source's DOA without threshold until the termination criterion is met. Unlike threshold-based algorithms, ISSL designs an active source detector network based on binary classifier to accept residual spatial spectrum and decide whether to stop the iteration. By doing so, our ISSL can deal with an arbitrary number of sources, even more than the number of sources seen during the training stage. The experimental results show that our ISSL achieves significant performance improvements in both DOA estimation and source number detection compared with the existing threshold-based algorithms.

Physically Consistent Learning of Conservative Lagrangian Systems with Gaussian Processes

This paper proposes a physically consistent Gaussian Process (GP) enabling the identification of uncertain Lagrangian systems. The function space is tailored according to the energy components of the Lagrangian and the differential equation structure, analytically guaranteeing physical and mathematical properties such as energy conservation and quadratic form. The novel formulation of Cholesky decomposed matrix kernels allow the probabilistic preservation of positive definiteness. Only differential input-to-output measurements of the function map are required while Gaussian noise is permitted in torques, velocities, and accelerations. We demonstrate the effectiveness of the approach in numerical simulation.

Using Autoencoders on Differentially Private Federated Learning GANs

Machine learning has been applied to almost all fields of computer science over the past decades. The introduction of GANs allowed for new possibilities in fields of medical research and text prediction. However, these new fields work with ever more privacy-sensitive data. In order to maintain user privacy, a combination of federated learning, differential privacy and GANs can be used to work with private data without giving away a users' privacy. Recently, two implementations of such combinations have been published: DP-Fed-Avg GAN and GS-WGAN. This paper compares their performance and introduces an alternative version of DP-Fed-Avg GAN that makes use of denoising techniques to combat the loss in accuracy that generally occurs when applying differential privacy and federated learning to GANs. We also compare the novel adaptation of denoised DP-Fed-Avg GAN to the state-of-the-art implementations in this field.

Data-driven reduced order models using invariant foliations, manifolds and autoencoders

This paper explores the question: how to identify a reduced order model from data. There are three ways to relate data to a model: invariant foliations, invariant manifolds and autoencoders. Invariant manifolds cannot be fitted to data unless a hardware in a loop system is used. Autoencoders only identify the portion of the phase space where the data is, which is not necessarily an invariant manifold. Therefore for off-line data the only option is an invariant foliation. We note that Koopman eigenfunctions also define invariant foliations, but they are limited by the assumption of linearity and resulting singularites. Finding an invariant foliation requires approximating high-dimensional functions. We propose two solutions. If an accurate reduced order model is sought, a sparse polynomial approximation is used, with polynomial coefficients that are sparse hierarchical tensors. If an invariant manifold is sought, as a leaf of a foliation, the required high-dimensional function can be approximated by a low-dimensional polynomial. The two methods can be combined to find an accurate reduced order model and an invariant manifold. We also analyse the reduced order model in case of a focus type equilibrium, typical in mechanical systems. We note that the nonlinear coordinate system defined by the invariant foliation and the invariant manifold distorts instantaneous frequencies and damping ratios, which we correct. Through examples we illustrate the calculation of invariant foliations and manifolds, and at the same time show that Koopman eigenfunctions and autoencoders fail to capture accurate reduced order models under the same conditions.

Reinforcement learning based adaptive metaheuristics

Parameter adaptation, that is the capability to automatically adjust an algorithm's hyperparameters depending on the problem being faced, is one of the main trends in evolutionary computation applied to numerical optimization. While several handcrafted adaptation policies have been proposed over the years to address this problem, only few attempts have been done so far at applying machine learning to learn such policies. Here, we introduce a general-purpose framework for performing parameter adaptation in continuous-domain metaheuristics based on state-of-the-art reinforcement learning algorithms. We demonstrate the applicability of this framework on two algorithms, namely Covariance Matrix Adaptation Evolution Strategies (CMA-ES) and Differential Evolution (DE), for which we learn, respectively, adaptation policies for the step-size (for CMA-ES), and the scale factor and crossover rate (for DE). We train these policies on a set of 46 benchmark functions at different dimensionalities, with various inputs to the policies, in two settings: one policy per function, and one global policy for all functions. Compared, respectively, to the Cumulative Step-size Adaptation (CSA) policy and to two well-known adaptive DE variants (iDE and jDE), our policies are able to produce competitive results in the majority of cases, especially in the case of DE.

Adversarial Robustness of Deep Neural Networks: A Survey from a Formal Verification Perspective

Neural networks have been widely applied in security applications such as spam and phishing detection, intrusion prevention, and malware detection. This black-box method, however, often has uncertainty and poor explainability in applications. Furthermore, neural networks themselves are often vulnerable to adversarial attacks. For those reasons, there is a high demand for trustworthy and rigorous methods to verify the robustness of neural network models. Adversarial robustness, which concerns the reliability of a neural network when dealing with maliciously manipulated inputs, is one of the hottest topics in security and machine learning. In this work, we survey existing literature in adversarial robustness verification for neural networks and collect 39 diversified research works across machine learning, security, and software engineering domains. We systematically analyze their approaches, including how robustness is formulated, what verification techniques are used, and the strengths and limitations of each technique. We provide a taxonomy from a formal verification perspective for a comprehensive understanding of this topic. We classify the existing techniques based on property specification, problem reduction, and reasoning strategies. We also demonstrate representative techniques that have been applied in existing studies with a sample model. Finally, we discuss open questions for future research.

MPClan: Protocol Suite for Privacy-Conscious Computations

The growing volumes of data being collected and its analysis to provide better services are creating worries about digital privacy. To address privacy concerns and give practical solutions, the literature has relied on secure multiparty computation. However, recent research has mostly focused on the small-party honest-majority setting of up to four parties, noting efficiency concerns. In this work, we extend the strategies to support a larger number of participants in an honest-majority setting with efficiency at the center stage. Cast in the preprocessing paradigm, our semi-honest protocol improves the online complexity of the decade-old state-of-the-art protocol of Damg\aa rd and Nielson (CRYPTO'07). In addition to having an improved online communication cost, we can shut down almost half of the parties in the online phase, thereby saving up to 50% in the system's operational costs. Our maliciously secure protocol also enjoys similar benefits and requires only half of the parties, except for one-time verification, towards the end. To showcase the practicality of the designed protocols, we benchmark popular applications such as deep neural networks, graph neural networks, genome sequence matching, and biometric matching using prototype implementations. Our improved protocols aid in bringing up to 60-80% savings in monetary cost over prior work.

Computational Complexity Evaluation of Neural Network Applications in Signal Processing

In this paper, we provide a systematic approach for assessing and comparing the computational complexity of neural network layers in digital signal processing. We provide and link four software-to-hardware complexity measures, defining how the different complexity metrics relate to the layers' hyper-parameters. This paper explains how to compute these four metrics for feed-forward and recurrent layers, and defines in which case we ought to use a particular metric depending on whether we characterize a more soft- or hardware-oriented application. One of the four metrics, called `the number of additions and bit shifts (NABS)', is newly introduced for heterogeneous quantization. NABS characterizes the impact of not only the bitwidth used in the operation but also the type of quantization used in the arithmetical operations. We intend this work to serve as a baseline for the different levels (purposes) of complexity estimation related to the neural networks' application in real-time digital signal processing, aiming at unifying the computational complexity estimation.

SECLEDS: Sequence Clustering in Evolving Data Streams via Multiple Medoids and Medoid Voting

Sequence clustering in a streaming environment is challenging because it is computationally expensive, and the sequences may evolve over time. K-medoids or Partitioning Around Medoids (PAM) is commonly used to cluster sequences since it supports alignment-based distances, and the k-centers being actual data items helps with cluster interpretability. However, offline k-medoids has no support for concept drift, while also being prohibitively expensive for clustering data streams. We therefore propose SECLEDS, a streaming variant of the k-medoids algorithm with constant memory footprint. SECLEDS has two unique properties: i) it uses multiple medoids per cluster, producing stable high-quality clusters, and ii) it handles concept drift using an intuitive Medoid Voting scheme for approximating cluster distances. Unlike existing adaptive algorithms that create new clusters for new concepts, SECLEDS follows a fundamentally different approach, where the clusters themselves evolve with an evolving stream. Using real and synthetic datasets, we empirically demonstrate that SECLEDS produces high-quality clusters regardless of drift, stream size, data dimensionality, and number of clusters. We compare against three popular stream and batch clustering algorithms. The state-of-the-art BanditPAM is used as an offline benchmark. SECLEDS achieves comparable F1 score to BanditPAM while reducing the number of required distance computations by 83.7%. Importantly, SECLEDS outperforms all baselines by 138.7% when the stream contains drift. We also cluster real network traffic, and provide evidence that SECLEDS can support network bandwidths of up to 1.08 Gbps while using the (expensive) dynamic time warping distance.

Dynamic network congestion pricing based on deep reinforcement learning

Traffic congestion is a serious problem in urban areas. Dynamic congestion pricing is one of the useful schemes to eliminate traffic congestion in strategic scale. However, in the reality, an optimal dynamic congestion pricing is very difficult or impossible to determine theoretically, because road networks are usually large and complicated, and behavior of road users is uncertain. To account for this challenge, this work proposes a dynamic congestion pricing method using deep reinforcement learning (DRL). It is designed to eliminate traffic congestion based on observable data in general large-scale road networks, by leveraging the data-driven nature of deep reinforcement learning. One of the novel elements of the proposed method is the distributed and cooperative learning scheme. Specifically, the DRL is implemented by a spatial-temporally distributed manner, and cooperation among DRL agents is established by novel techniques we call spatially shared reward and temporally switching learning. It enables fast and computationally efficient learning in large-scale networks. The numerical experiments using Sioux Falls Network showed that the proposed method works well thanks to the novel learning scheme.

ML-Based Approach for NFL Defensive Pass Interference Prediction Using GPS Tracking Data

Defensive Pass Interference (DPI) is one of the most impactful penalties in the NFL. DPI is a spot foul, yielding an automatic first down to the team in possession. With such an influence on the game, referees have no room for a mistake. It is also a very rare event, which happens 1-2 times per 100 pass attempts. With technology improving and many IoT wearables being put on the athletes to collect valuable data, there is a solid ground for applying machine learning (ML) techniques to improve every aspect of the game. The work presented here is the first attempt in predicting DPI using player tracking GPS data. The data we used was collected by NFL's Next Gen Stats throughout the 2018 regular season. We present ML models for highly imbalanced time-series binary classification: LSTM, GRU, ANN, and Multivariate LSTM-FCN. Results showed that using GPS tracking data to predict DPI has limited success. The best performing models had high recall with low precision which resulted in the classification of many false positive examples. Looking closely at the data confirmed that there is just not enough information to determine whether a foul was committed. This study might serve as a filter for multi-step pipeline for video sequence classification which could be able to solve this problem.

"You Can't Fix What You Can't Measure": Privately Measuring Demographic Performance Disparities in Federated Learning

Federated learning allows many devices to collaborate in the training of machine learning models. As in traditional machine learning, there is a growing concern that models trained with federated learning may exhibit disparate performance for different demographic groups. Existing solutions to measure and ensure equal model performance across groups require access to information about group membership, but this access is not always available or desirable, especially under the privacy aspirations of federated learning. We study the feasibility of measuring such performance disparities while protecting the privacy of the user's group membership and the federated model's performance on the user's data. Protecting both is essential for privacy, because they may be correlated, and thus learning one may reveal the other. On the other hand, from the utility perspective, the privacy-preserved data should maintain the correlation to ensure the ability to perform accurate measurements of the performance disparity. We achieve both of these goals by developing locally differentially private mechanisms that preserve the correlations between group membership and model performance. To analyze the effectiveness of the mechanisms, we bound their error in estimating the disparity when optimized for a given privacy budget, and validate these bounds on synthetic data. Our results show that the error rapidly decreases for realistic numbers of participating clients, demonstrating that, contrary to what prior work suggested, protecting the privacy of protected attributes is not necessarily in conflict with identifying disparities in the performance of federated models.

Towards FPGA Implementation of Neural Network-Based Nonlinearity Mitigation Equalizers in Coherent Optical Transmission Systems

For the first time, recurrent and feedforward neural network-based equalizers for nonlinearity compensation are implemented in an FPGA, with a level of complexity comparable to that of a dispersion equalizer. We demonstrate that the NN-based equalizers can outperform a 1 step-per-span DBP.

MULTI-FLGANs: Multi-Distributed Adversarial Networks for Non-IID distribution

Federated learning is an emerging concept in the domain of distributed machine learning. This concept has enabled GANs to benefit from the rich distributed training data while preserving privacy. However, in a non-iid setting, current federated GAN architectures are unstable, struggling to learn the distinct features and vulnerable to mode collapse. In this paper, we propose a novel architecture MULTI-FLGAN to solve the problem of low-quality images, mode collapse and instability for non-iid datasets. Our results show that MULTI-FLGAN is four times as stable and performant (i.e. high inception score) on average over 20 clients compared to baseline FLGAN.

SCAI: A Spectral data Classification framework with Adaptive Inference for the IoT platform

Currently, it is a hot research topic to realize accurate, efficient, and real-time identification of massive spectral data with the help of deep learning and IoT technology. Deep neural networks played a key role in spectral analysis. However, the inference of deeper models is performed in a static manner, and cannot be adjusted according to the device. Not all samples need to allocate all computation to reach confident prediction, which hinders maximizing the overall performance. To address the above issues, we propose a Spectral data Classification framework with Adaptive Inference. Specifically, to allocate different computations for different samples while better exploiting the collaboration among different devices, we leverage Early-exit architecture, place intermediate classifiers at different depths of the architecture, and the model outputs the results when the prediction confidence reaches a preset threshold. We propose a training paradigm of self-distillation learning, the deepest classifier performs soft supervision on the shallow ones to maximize their performance and training speed. At the same time, to mitigate the vulnerability of performance to the location and number settings of intermediate classifiers in the Early-exit paradigm, we propose a Position-Adaptive residual network. It can adjust the number of layers in each block at different curve positions, so it can focus on important positions of the curve (e.g.: Raman peak), and accurately allocate the appropriate computational budget based on task performance and computing resources. To the best of our knowledge, this paper is the first attempt to conduct optimization by adaptive inference for spectral detection under the IoT platform. We conducted many experiments, the experimental results show that our proposed method can achieve higher performance with less computational budget than existing methods.

AdAUC: End-to-end Adversarial AUC Optimization Against Long-tail Problems

It is well-known that deep learning models are vulnerable to adversarial examples. Existing studies of adversarial training have made great progress against this challenge. As a typical trait, they often assume that the class distribution is overall balanced. However, long-tail datasets are ubiquitous in a wide spectrum of applications, where the amount of head class instances is larger than the tail classes. Under such a scenario, AUC is a much more reasonable metric than accuracy since it is insensitive toward class distribution. Motivated by this, we present an early trial to explore adversarial training methods to optimize AUC. The main challenge lies in that the positive and negative examples are tightly coupled in the objective function. As a direct result, one cannot generate adversarial examples without a full scan of the dataset. To address this issue, based on a concavity regularization scheme, we reformulate the AUC optimization problem as a saddle point problem, where the objective becomes an instance-wise function. This leads to an end-to-end training protocol. Furthermore, we provide a convergence guarantee of the proposed algorithm. Our analysis differs from the existing studies since the algorithm is asked to generate adversarial examples by calculating the gradient of a min-max problem. Finally, the extensive experimental results show the performance and robustness of our algorithm in three long-tail datasets.

Neural Networks with A La Carte Selection of Activation Functions

Activation functions (AFs), which are pivotal to the success (or failure) of a neural network, have received increased attention in recent years, with researchers seeking to design novel AFs that improve some aspect of network performance. In this paper we take another direction, wherein we combine a slew of known AFs into successful architectures, proposing three methods to do so beneficially: 1) generate AF architectures at random, 2) use Optuna, an automatic hyper-parameter optimization software framework, with a Tree-structured Parzen Estimator (TPE) sampler, and 3) use Optuna with a Covariance Matrix Adaptation Evolution Strategy (CMA-ES) sampler. We show that all methods often produce significantly better results for 25 classification problems when compared with a standard network composed of ReLU hidden units and a softmax output unit. Optuna with the TPE sampler emerged as the best AF architecture-producing method.

MultiSAGE: a multiplex embedding algorithm for inter-layer link prediction

Research on graph representation learning has received great attention in recent years. However, most of the studies so far have focused on the embedding of single-layer graphs. The few studies dealing with the problem of representation learning of multilayer structures rely on the strong hypothesis that the inter-layer links are known, and this limits the range of possible applications. Here we propose MultiSAGE, a generalization of the GraphSAGE algorithm that allows to embed multiplex networks. We show that MultiSAGE is capable to reconstruct both the intra-layer and the inter-layer connectivity, outperforming GraphSAGE, which has been designed for simple graphs. Next, through a comprehensive experimental analysis, we shed light also on the performance of the embedding, both in simple and in multiplex networks, showing that either the density of the graph or the randomness of the links strongly influences the quality of the embedding.

Modeling Adaptive Platoon and Reservation Based Autonomous Intersection Control: A Deep Reinforcement Learning Approach

As a strategy to reduce travel delay and enhance energy efficiency, platooning of connected and autonomous vehicles (CAVs) at non-signalized intersections has become increasingly popular in academia. However, few studies have attempted to model the relation between the optimal platoon size and the traffic conditions around the intersection. To this end, this study proposes an adaptive platoon based autonomous intersection control model powered by deep reinforcement learning (DRL) technique. The model framework has following two levels: the first level adopts a First Come First Serve (FCFS) reservation based policy integrated with a nonconflicting lane selection mechanism to determine vehicles' passing priority; and the second level applies a deep Q-network algorithm to identify the optimal platoon size based on the real-time traffic condition of an intersection. When being tested on a traffic micro-simulator, our proposed model exhibits superior performances on travel efficiency and fuel conservation as compared to the state-of-the-art methods.

FLVoogd: Robust And Privacy Preserving Federated Learning

In this work, we propose FLVoogd, an updated federated learning method in which servers and clients collaboratively eliminate Byzantine attacks while preserving privacy. In particular, servers use automatic Density-based Spatial Clustering of Applications with Noise (DBSCAN) combined with S2PC to cluster the benign majority without acquiring sensitive personal information. Meanwhile, clients build dual models and perform test-based distance controlling to adjust their local models toward the global one to achieve personalizing. Our framework is automatic and adaptive that servers/clients don't need to tune the parameters during the training. In addition, our framework leverages Secure Multi-party Computation (SMPC) operations, including multiplications, additions, and comparison, where costly operations, like division and square root, are not required. Evaluations are carried out on some conventional datasets from the image classification field. The result shows that FLVoogd can effectively reject malicious uploads in most scenarios; meanwhile, it avoids data leakage from the server-side.

Data-driven discovery of novel 2D materials by deep generative models

Efficient algorithms to generate candidate crystal structures with good stability properties can play a key role in data-driven materials discovery. Here we show that a crystal diffusion variational autoencoder (CDVAE) is capable of generating two-dimensional (2D) materials of high chemical and structural diversity and formation energies mirroring the training structures. Specifically, we train the CDVAE on 2615 2D materials with energy above the convex hull $\Delta H_{\mathrm{hull}}< 0.3$ eV/atom, and generate 5003 materials that we relax using density functional theory (DFT). We also generate 14192 new crystals by systematic element substitution of the training structures. We find that the generative model and lattice decoration approach are complementary and yield materials with similar stability properties but very different crystal structures and chemical compositions. In total we find 11630 predicted new 2D materials, where 8599 of these have $\Delta H_{\mathrm{hull}}< 0.3$ eV/atom as the seed structures, while 2004 are within 50 meV of the convex hull and could potentially be synthesized. The relaxed atomic structures of all the materials are available in the open Computational 2D Materials Database (C2DB). Our work establishes the CDVAE as an efficient and reliable crystal generation machine, and significantly expands the space of 2D materials.

Geometry Interaction Knowledge Graph Embeddings

Knowledge graph (KG) embeddings have shown great power in learning representations of entities and relations for link prediction tasks. Previous work usually embeds KGs into a single geometric space such as Euclidean space (zero curved), hyperbolic space (negatively curved) or hyperspherical space (positively curved) to maintain their specific geometric structures (e.g., chain, hierarchy and ring structures). However, the topological structure of KGs appears to be complicated, since it may contain multiple types of geometric structures simultaneously. Therefore, embedding KGs in a single space, no matter the Euclidean space, hyperbolic space or hyperspheric space, cannot capture the complex structures of KGs accurately. To overcome this challenge, we propose Geometry Interaction knowledge graph Embeddings (GIE), which learns spatial structures interactively between the Euclidean, hyperbolic and hyperspherical spaces. Theoretically, our proposed GIE can capture a richer set of relational information, model key inference patterns, and enable expressive semantic matching across entities. Experimental results on three well-established knowledge graph completion benchmarks show that our GIE achieves the state-of-the-art performance with fewer parameters.

Multi-Agent Deep Reinforcement Learning for Cost- and Delay-Sensitive Virtual Network Function Placement and Routing

This paper proposes an effective and novel multiagent deep reinforcement learning (MADRL)-based method for solving the joint virtual network function (VNF) placement and routing (P&R), where multiple service requests with differentiated demands are delivered at the same time. The differentiated demands of the service requests are reflected by their delay- and cost-sensitive factors. We first construct a VNF P&R problem to jointly minimize a weighted sum of service delay and resource consumption cost, which is NP-complete. Then, the joint VNF P&R problem is decoupled into two iterative subtasks: placement subtask and routing subtask. Each subtask consists of multiple concurrent parallel sequential decision processes. By invoking the deep deterministic policy gradient method and multi-agent technique, an MADRL-P&R framework is designed to perform the two subtasks. The new joint reward and internal rewards mechanism is proposed to match the goals and constraints of the placement and routing subtasks. We also propose the parameter migration-based model-retraining method to deal with changing network topologies. Corroborated by experiments, the proposed MADRL-P&R framework is superior to its alternatives in terms of service cost and delay, and offers higher flexibility for personalized service demands. The parameter migration-based model-retraining method can efficiently accelerate convergence under moderate network topology changes.

Competence-based Multimodal Curriculum Learning for Medical Report Generation

Medical report generation task, which targets to produce long and coherent descriptions of medical images, has attracted growing research interests recently. Different from the general image captioning tasks, medical report generation is more challenging for data-driven neural models. This is mainly due to 1) the serious data bias and 2) the limited medical data. To alleviate the data bias and make best use of available data, we propose a Competence-based Multimodal Curriculum Learning framework (CMCL). Specifically, CMCL simulates the learning process of radiologists and optimizes the model in a step by step manner. Firstly, CMCL estimates the difficulty of each training instance and evaluates the competence of current model; Secondly, CMCL selects the most suitable batch of training instances considering current model competence. By iterating above two steps, CMCL can gradually improve the model's performance. The experiments on the public IU-Xray and MIMIC-CXR datasets show that CMCL can be incorporated into existing models to improve their performance.