Datasets:

debajyotidasgupta

/

dblp_2022_2023

like 0

Title: Minimum degree thresholds for Hamilton (k/2)-cycles in k-uniform hypergraphs Abstract: For any even integer k≥6, integer d such that k/2≤d≤k−1, and sufficiently large n∈(k/2)N, we find a tight minimum d-degree condition that guarantees the existence of a Hamilton (k/2)-cycle in every k-uniform hypergraph on n vertices. When n∈kN, the degree condition coincides with the one for the existence of perfect matchings provided by Rödl, Ruciński and Szemerédi (for d=k−1) and Treglown and Zhao (for d≥k/2), and thus our result strengthens theirs in this case.

Title: Visual Camera Re-Localization From RGB and RGB-D Images Using DSAC Abstract: We describe a learning-based system that estimates the camera position and orientation from a single input image relative to a known environment. The system is flexible w.r.t. the amount of information available at test and at training time, catering to different applications. Input images can be RGB-D or RGB, and a 3D model of the environment can be utilized for training but is not necessary. In the minimal case, our system requires only RGB images and ground truth poses at training time, and it requires only a single RGB image at test time. The framework consists of a deep neural network and fully differentiable pose optimization. The neural network predicts so called scene coordinates, i.e., dense correspondences between the input image and 3D scene space of the environment. The pose optimization implements robust fitting of pose parameters using differentiable RANSAC ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DSAC</i> ) to facilitate end-to-end training. The system, an extension of DSAC++ and referred to as DSAC*, achieves state-of-the-art accuracy on various public datasets for RGB-based re-localization, and competitive accuracy for RGB-D based re-localization.

Title: Security Measures for Grids Against Rank-1 Undetectable Time-Synchronization Attacks Abstract: Time-synchronization attacks on phasor measurement units (PMUs) pose a real threat to smart grids; it was shown that they are feasible in practice and that they can have a nonnegligible negative impact on state estimation, without triggering the bad data detection mechanisms. Previous works identified vulnerability conditions when targeted PMUs measure a single phasor. Yet, PMUs are capable of measuring several quantities. We present novel vulnerability conditions in the general case, where PMUs measure any number of phasors and can share the same time reference. One is a sufficient condition that does not depend on the measurement values. We propose a security requirement that prevents it and provide a greedy offline algorithm that enforces it. If this security requirement is satisfied, there is still a possibility that the grid can be attacked, although we conjecture that it is very unlikely. We identify two sufficient and necessary vulnerability conditions, which depend on the measurement values. For each, we provide a metric that shows the distance between the observed and vulnerability conditions. We recommend their monitoring for security. Numerical results on the IEEE-39 bus benchmark with real load profiles show that the measurements of a grid satisfying our security requirement are far from vulnerable.

Title: Weighted least squares estimators for the Parzen tail index Abstract: Estimation of the tail index of heavy-tailed distributions and its applications are essential in many research areas. We propose a class of weighted least squares (WLS) estimators for the Parzen tail index. Our approach is based on the method developed by Holan and McElroy (J Stat Plan Inference 140(12):3693–3708, 2010). We investigate consistency and asymptotic normality of the WLS estimators. Through a simulation study, we make a comparison with the Hill, Pickands, DEdH (Dekkers, Einmahl and de Haan) and ordinary least squares (OLS) estimators using the mean square error as criterion. The results show that in a restricted model some members of the WLS estimators are competitive with the Pickands, DEdH and OLS estimators.

Title: On standard quadratic programs with exact and inexact doubly nonnegative relaxations Abstract: The problem of minimizing a (nonconvex) quadratic form over the unit simplex, referred to as a standard quadratic program, admits an exact convex conic formulation over the computationally intractable cone of completely positive matrices. Replacing the intractable cone in this formulation by the larger but tractable cone of doubly nonnegative matrices, i.e., the cone of positive semidefinite and componentwise nonnegative matrices, one obtains the so-called doubly nonnegative relaxation, whose optimal value yields a lower bound on that of the original problem. We present a full algebraic characterization of the set of instances of standard quadratic programs that admit an exact doubly nonnegative relaxation. This characterization yields an algorithmic recipe for constructing such an instance. In addition, we explicitly identify three families of instances for which the doubly nonnegative relaxation is exact. We establish several relations between the so-called convexity graph of an instance and the tightness of the doubly nonnegative relaxation. We also provide an algebraic characterization of the set of instances for which the doubly nonnegative relaxation has a positive gap and show how to construct such an instance using this characterization.

Title: Tuning as convex optimisation: a polynomial tuner for multi-parametric combinatorial samplers. Abstract: Combinatorial samplers are algorithmic schemes devised for the approximate- and exact-size generation of large random combinatorial structures, such as context-free words, various tree-like data structures, maps, tilings, or even RNA sequences. In their multi-parametric variants, combinatorial samplers are adapted to combinatorial specifications with additional parameters, allowing for a more flexible control over the output profile of parametrised combinatorial patterns. One can control, for instance, the number of leaves, profile of node degrees in trees or the number of certain sub-patterns in generated strings. However, such a flexible control requires an additional and nontrivial tuning procedure. Using techniques of convex optimisation, we present an efficient polynomial tuning algorithm for multi-parametric combinatorial specifications. For a given combinatorial system of description length $L$ with $d$ tuning parameters and target size parameter value $n$, our algorithm runs in time $O(d^{3.5} L \log n)$. We demonstrate the effectiveness of our method on a series of practical examples, including rational, algebraic, and so-called P\'olya specifications. We show how our method can be adapted to a broad range of less typical combinatorial constructions, including symmetric polynomials, labelled sets and cycles with cardinality lower bounds, simple increasing trees or substitutions. Finally, we discuss some practical aspects of our prototype tuner implementation and provide its benchmark results.

Title: Implicit algorithms for eigenvector nonlinearities Abstract: We study and derive algorithms for nonlinear eigenvalue problems, where the system matrix depends on the eigenvector, or several eigenvectors (or their corresponding invariant subspace). The algorithms are derived from an implicit viewpoint. More precisely, we change the Newton update equation in a way that the next iterate does not only appear linearly in the update equation. Although the modifications of the update equation make the methods implicit, we show how corresponding iterates can be computed explicitly. Therefore, we can carry out steps of the implicit method using explicit procedures. In several cases, these procedures involve a solution of standard eigenvalue problems. We propose two modifications, one of the modifications leads directly to a well-established method (the self-consistent field iteration) whereas the other method is to our knowledge new and has several attractive properties. Convergence theory is provided along with several simulations which illustrate the properties of the algorithms.

Title: Federated Over-Air Subspace Tracking From Incomplete and Corrupted Data Abstract: In this work we study the problem of Subspace Tracking with missing data (ST-miss) and outliers (Robust ST-miss). We propose a novel algorithm, and provide a guarantee for both these problems. Unlike past work on this topic, the current work does not impose the piecewise constant subspace change assumption. Additionally, the proposed algorithm is much simpler (uses fewer parameters) than our previous work. Secondly, we extend our approach and its analysis to provably solving these problems when the data is federated and when the over-air data communication modality is used for information exchange between the K peer nodes and the center. We validate our theoretical claims with extensive numerical experiments.

Title: Semantic Edge Detection with Diverse Deep Supervision Abstract: Semantic edge detection (SED), which aims at jointly extracting edges as well as their category information, has far-reaching applications in domains such as semantic segmentation, object proposal generation, and object recognition. SED naturally requires achieving two distinct supervision targets: locating fine detailed edges and identifying high-level semantics. Our motivation comes from the hypothesis that such distinct targets prevent state-of-the-art SED methods from effectively using deep supervision to improve results. To this end, we propose a novel fully convolutional neural network using diverse deep supervision within a multi-task framework where bottom layers aim at generating category-agnostic edges, while top layers are responsible for the detection of category-aware semantic edges. To overcome the hypothesized supervision challenge, a novel information converter unit is introduced, whose effectiveness has been extensively evaluated on SBD and Cityscapes datasets.

Title: Automated regression unit test generation for program merges Abstract: We propose test oracles for real-world program merges including two-way, three-way, and octopus merges. On this basis, we implemented a tool called TOM to automatically generate test cases to reveal merge conflicts. In addition, we designed the benchmark MCon4J to support further studies on merges. In our experiments, a total of 45 conflict three-way merges and 87 conflict octopus merges were detected using TOM, while the verification-based tool SafeMerge failed to work on MCon4J.

Title: Robustness of Uncertain Switching Nonlinear Feedback Systems Against Large Time-Variation Abstract: For a multiple-input–multiple-output (MIMO) nonlinear feedback system, the robustness against uncertain time-variation (slow or infrequently large) in the feedback loop is investigated in an input–output framework. A couple of sufficient conditions in terms of bounds on average rates of time-variation for the system to be stable are derived. The conditions provide a useful tool for designing adaptive controllers for systems against uncertain large time-variation.

Title: Learning to Compare Relation: Semantic Alignment for Few-Shot Learning Abstract: Few-shot learning is a fundamental and challenging problem since it requires recognizing novel categories from only a few examples. The objects for recognition have multiple variants and can locate anywhere in images. Directly comparing query images with example images can not handle content misalignment. The representation and metric for comparison are critical but challenging to learn due to the scarcity and wide variation of the samples in few-shot learning. In this paper, we present a novel semantic alignment model to compare relations, which is robust to content misalignment. We propose to add two key ingredients to existing few-shot learning frameworks for better feature and metric learning ability. First, we introduce a semantic alignment loss to align the relation statistics of the features from samples that belong to the same category. And second, local and global mutual information maximization is introduced, allowing for representations that contain locally-consistent and intra-class shared information across structural locations in an image. Furthermore, we introduce a principled approach to weigh multiple loss functions by considering the homoscedastic uncertainty of each stream. We conduct extensive experiments on several few-shot learning datasets. Experimental results show that the proposed method is capable of comparing relations with semantic alignment strategies, and achieves state-of-the-art performance.

Title: Dimension-free convergence rates for gradient Langevin dynamics in RKHS. Abstract: Gradient Langevin dynamics (GLD) and stochastic GLD (SGLD) have attracted considerable attention lately, as a way to provide convergence guarantees in a non-convex setting. However, the known rates grow exponentially with the dimension of the space under the dissipative condition. In this work, we provide a convergence analysis of GLD and SGLD when the optimization space is an infinite-dimensional Hilbert space. More precisely, we derive non-asymptotic, dimension-free convergence rates for GLD/SGLD when performing regularized non-convex optimization in a reproducing kernel Hilbert space. Amongst others, the convergence analysis relies on the properties of a stochastic differential equation, its discrete time Galerkin approximation and the geometric ergodicity of the associated Markov chains.

Title: Statistical power for cluster analysis Abstract: Cluster algorithms are gaining in popularity in biomedical research due to their compelling ability to identify discrete subgroups in data, and their increasing accessibility in mainstream software. While guidelines exist for algorithm selection and outcome evaluation, there are no firmly established ways of computing a priori statistical power for cluster analysis. Here, we estimated power and classification accuracy for common analysis pipelines through simulation. We systematically varied subgroup size, number, separation (effect size), and covariance structure. We then subjected generated datasets to dimensionality reduction approaches (none, multi-dimensional scaling, or uniform manifold approximation and projection) and cluster algorithms (k-means, agglomerative hierarchical clustering with Ward or average linkage and Euclidean or cosine distance, HDBSCAN). Finally, we directly compared the statistical power of discrete (k-means), “fuzzy” (c-means), and finite mixture modelling approaches (which include latent class analysis and latent profile analysis). We found that clustering outcomes were driven by large effect sizes or the accumulation of many smaller effects across features, and were mostly unaffected by differences in covariance structure. Sufficient statistical power was achieved with relatively small samples (N = 20 per subgroup), provided cluster separation is large (Δ = 4). Finally, we demonstrated that fuzzy clustering can provide a more parsimonious and powerful alternative for identifying separable multivariate normal distributions, particularly those with slightly lower centroid separation (Δ = 3). Traditional intuitions about statistical power only partially apply to cluster analysis: increasing the number of participants above a sufficient sample size did not improve power, but effect size was crucial. Notably, for the popular dimensionality reduction and clustering algorithms tested here, power was only satisfactory for relatively large effect sizes (clear separation between subgroups). Fuzzy clustering provided higher power in multivariate normal distributions. Overall, we recommend that researchers (1) only apply cluster analysis when large subgroup separation is expected, (2) aim for sample sizes of N = 20 to N = 30 per expected subgroup, (3) use multi-dimensional scaling to improve cluster separation, and (4) use fuzzy clustering or mixture modelling approaches that are more powerful and more parsimonious with partially overlapping multivariate normal distributions.

Title: New Constructions Related to the Polynomial Sphere Recognition Problem Abstract: We construct a simply connected 2-complex C embeddable in 3-space such that for any embedding of C in  $${\mathbb {S}}^3$$ , any edge contraction forms a minor of the 2-complex not embeddable in 3-space. Additionally, we construct the 2-complex C such that every edge of C forms a nontrivial knot in any of the embeddings of C in  $${\mathbb {S}}^3$$ . This answers a question of Saul Schleimer.

Title: Weak texture information map guided image super-resolution with deep residual networks Abstract: Limited by the poor quality of the camera, transmission bandwidth, excessive compression and other factors, low-resolution images widely exist in our lives. Single image super-resolution method is a kind of image processing task which can obtain high-resolution image from corresponding source low-resolution image. With the development of deep learning technology, a series of deep learning methods have brought crucial improvement for SISR problem. However, we observe that no matter how deep the network structures for SISR are designed, they usually have poor performances on restoration of tiny or irregular details, we call them weak texture information. The main reason for this phenomenon is that weak texture information is not obvious relative to the salient features, so as weak texture information feature is less extracted in the process of neural network feature extraction. To address this problem, we propose a SISR method which owns unique weak texture information prediction module and call it as weak texture information map guided image super-resolution with deep residual networks. In our network structure, two auxiliary sub-networks work together to capture grab details information and predict weak texture information. Then, predicted weak texture information will be fused into main sub-network. Therefore, our network can obtain more irregular details which usually miss in deep learning based methods. Finally, both qualitative experiments and visual effects demonstrate the effectiveness of our proposed algorithm. Specifically, our method can restore more irregular image details.

Title: Deep attention aware feature learning for person re-Identification Abstract: •We propose to learn global and local attention aware features for person ReID.•Two additional branches are introduced to realize the proposed attention aware feature learning in the training stage, and they are removed in the inference time to keep the same model size and inference speed.•Ablation studies and visualization results are included to help understanding the proposed method.•Significant performance improvements over existing methods are achieved on five widely used benchmarks.

Title: The Maximum-Level Vertex in an Arrangement of Lines Abstract: Let L be a set of n lines in the plane, not necessarily in general position. We present an efficient algorithm for finding all the vertices of the arrangement A(L) of maximum level, where the level of a vertex v is the number of lines of L that pass strictly below v. The problem, posed in Exercise 8.13 in de Berg et al. (Computational Geometry. Algorithms and Applications. Springer, Berlin (2008)), appears to be much harder than it seems at first sight, as this vertex might not be on the upper envelope of the lines. We first assume that all the lines of L are distinct, and distinguish between two cases, depending on whether or not the upper envelope of L contains a bounded edge. In the former case, we show that the number of lines of L that pass above any maximum level vertex v(0) is only O (log n). In the latter case, we establish a similar property that holds after we remove some of the lines that are incident to the single vertex of the upper envelope. We present algorithms that run, in both cases, in optimal O (n log n) time. We then consider the case where the lines of L are not necessarily distinct. This setup is more challenging, and for this case we present an algorithm that computes all the maximum-level vertices in time O(n(4/3) log(3) n). Finally, we consider a related combinatorial question for degenerate arrangements, where many lines may intersect in a single point, but all the lines are distinct: We bound the complexity of the weighted k-level in such an arrangement, where the weight of a vertex is the number of lines that pass through the vertex. We show that the bound in this case is O(n(4/3)), which matches the corresponding bound for non-degenerate arrangements, and we use this bound in the analysis of one of our algorithms.

Title: Proportional Power-Sharing Control of Distributed Generators in Microgrids Abstract: This research addresses distributed proportional power-sharing of inverter-based distributed generators (DGs) in microgrids under variations in the maximum power capacity of DGs. A microgrid can include renewable energy resources such as wind turbines, solar panels, fuel cells, etc. The intermittent nature of such energy resources causes variations in their maximum power capacities. Since DGs in microgrids can be regarded as multiagent-systems (MASs), a consensus algorithm is designed to have the DGs to generate their output power in proportion to their maximum capacities under capacity fluctuations. A change in power capacity of a DG triggers the consensus algorithm, which uses a communication map at the cyber layer to estimate the corresponding change. The convergence rate of the algorithm is analytically established and bounds on allowable capacity fluctuations are derived based on practical constraints. During the transient time of reaching a consensus, the delivered power may not match the load power demand. To eliminate this mismatch, a control law is augmented that consists of a finite-time consensus algorithm embedded within the overarching power-sharing consensus algorithm. The effectiveness of the distributed controller is assessed through simulation of a microgrid consisting of a realistic model of inverter-based DGs. Details of the microgrid model, its controller structures, and a comprehensive list of parameters are provided.

Title: Federating recommendations using differentially private prototypes Abstract: Machine learning methods exploit similarities in users’ activity patterns to provide recommendations in applications across a wide range of fields including entertainment, dating, and commerce. However, in domains that demand protection of personally sensitive data, such as medicine or banking, how can we learn recommendation models without accessing the sensitive data and without inadvertently leaking private information? Many situations in the medical field prohibit centralizing the data from different hospitals and thus require learning from information kept in separate databases. We propose a new federated approach to learning global and local private models for recommendation without collecting raw data, user statistics, or information about personal preferences. Our method produces a set of locally learned prototypes that allow us to infer global behavioral patterns while providing differential privacy guarantees for users in any database of the system. By requiring only two rounds of communication, we both reduce the communication costs and avoid excessive privacy loss associated with typical federated learning iterative procedures. We test our framework on synthetic data, real federated medical data, and a federated version of Movielens ratings. We show that local adaptation of the global model allows the proposed method to outperform centralized matrix-factorization-based recommender system models, both in terms of the accuracy of matrix reconstruction and in terms of the relevance of recommendations, while maintaining provable privacy guarantees. We also show that our method is more robust and has smaller variance than individual models learned by independent entities.

Title: Uncertainty-Aware Variational-Recurrent Imputation Network for Clinical Time Series Abstract: Electronic health records (EHR) consist of longitudinal clinical observations portrayed with sparsity, irregularity, and high dimensionality, which become major obstacles in drawing reliable downstream clinical outcomes. Although there exist great numbers of imputation methods to tackle these issues, most of them ignore correlated features, temporal dynamics, and entirely set aside the uncertainty. Since the missing value estimates involve the risk of being inaccurate, it is appropriate for the method to handle the less certain information differently than the reliable data. In that regard, we can use the uncertainties in estimating the missing values as the fidelity score to be further utilized to alleviate the risk of biased missing value estimates. In this work, we propose a novel variational-recurrent imputation network, which unifies an imputation and a prediction network by taking into account the correlated features, temporal dynamics, as well as uncertainty. Specifically, we leverage the deep generative model in the imputation, which is based on the distribution among variables, and a recurrent imputation network to exploit the temporal relations, in conjunction with utilization of the uncertainty. We validated the effectiveness of our proposed model on two publicly available real-world EHR datasets: 1) PhysioNet Challenge 2012 and 2) MIMIC-III, and compared the results with other competing state-of-the-art methods in the literature.

Title: Fusing physics-based and deep learning models for prognostics Abstract: Physics-based and data-driven models for remaining useful lifetime (RUL) prediction typically suffer from two major challenges that limit their applicability to complex real-world domains: (1) the incompleteness of physics-based models and (2) the limited representativeness of the training dataset for data-driven models. Combining the advantages of these two approaches while overcoming some of their limitations, we propose a novel hybrid framework for fusing the information from physics-based performance models with deep learning algorithms for prognostics of complex safety-critical systems. In the proposed framework, we use physics-based performance models to infer unobservable model parameters related to a system’s components health by solving a calibration problem. These parameters are subsequently combined with sensor readings and used as input to a deep neural network, thereby generating a data-driven prognostics model with physics-augmented features. The performance of the hybrid framework is evaluated on an extensive case study comprising run-to-failure degradation trajectories from a fleet of nine turbofan engines under real flight conditions. The experimental results show that the hybrid framework outperforms purely data-driven approaches by extending the prediction horizon by nearly 127%. Furthermore, it requires less training data and is less sensitive to the limited representativeness of the dataset as compared to purely data-driven approaches. Furthermore, we demonstrated the feasibility of the proposed framework on the original CMAPSS dataset, thereby confirming its superior performance.

Title: Spatiotemporal Learning of Multivehicle Interaction Patterns in Lane-Change Scenarios Abstract: Interpretation of common-yet-challenging inter- action scenarios can benefit well-founded decisions for autonomous vehicles. Previous research achieved this using their prior knowledge of specific scenarios with predefined models, limiting their adaptive capabilities. This paper describes a Bayesian nonparametric approach that leverages continuous (i.e., Gaussian processes) and discrete (i.e., Dirichlet processes) stochastic processes to reveal underlying interaction patterns of the ego vehicle with other nearby vehicles. Our model relaxes dependency on the number of surrounding vehicles by developing an acceleration-sensitive velocity field based on Gaussian processes. The experiment results demonstrate that the velocity field can represent the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">spatial</i> interactions between the ego vehicle and its surroundings. A discrete Bayesian nonparametric model, integrating Dirichlet processes and hidden Markov models, is developed to learn the interaction patterns over the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">temporal</i> space by segmenting and clustering the sequential interaction data into interpretable granular patterns automatically. We then evaluate our approach in the highway discretionary lane-change scenarios using the highD dataset collected from real-world settings. Results demonstrate that our proposed Bayesian nonparametric approach provides an insight into the complicated discretionary lane-change interactions of the ego vehicle with multiple surrounding traffic participants based on the interpretable interaction patterns and their transition properties in temporal relationships. Our proposed approach sheds light on efficiently analyzing other kinds of multi-agent interactions, such as vehicle-pedestrian interactions. View associated demos via: <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><uri>https://chengyuan-zhang.github.io/Multivehicle-Interaction</uri></monospace> .

Title: Unsupervised Learning of Depth, Optical Flow and Pose With Occlusion From 3D Geometry Abstract: In autonomous driving, monocular sequences contain lots of information. Monocular depth estimation, camera ego-motion estimation and optical flow estimation in consecutive frames are high-profile concerns recently. By analyzing tasks above, pixels in the middle frame are modeled into three parts: the rigid region, the non-rigid region, and the occluded region. In joint unsupervised training of dep...

Title: AlignSeg: Feature-Aligned Segmentation Networks Abstract: Aggregating features in terms of different convolutional blocks or contextual embeddings has been proven to be an effective way to strengthen feature representations for semantic segmentation. However, most of the current popular network architectures tend to ignore the misalignment issues during the feature aggregation process caused by step-by-step downsampling operations and indiscriminate cont...

Title: Generating Higher-Fidelity Synthetic Datasets with Privacy Guarantees Abstract: We consider the problem of enhancing user privacy in common data analysis and machine learning development tasks, such as data annotation and inspection, by substituting the real data with samples from a generative adversarial network. We propose employing Bayesian differential privacy as the means to achieve a rigorous theoretical guarantee while providing a better privacy-utility trade-off. We demonstrate experimentally that our approach produces higher-fidelity samples compared to prior work, allowing to (1) detect more subtle data errors and biases, and (2) reduce the need for real data labelling by achieving high accuracy when training directly on artificial samples.

Title: Fast predictive uncertainty for classification with Bayesian deep networks. Abstract: In Bayesian Deep Learning, distributions over the output of classification neural networks are often approximated by first constructing a Gaussian distribution over the weights, then sampling from it to receive a distribution over the softmax outputs. This is costly. We reconsider old work (Laplace Bridge) to construct a Dirichlet approximation of this softmax output distribution, which yields an analytic map between Gaussian distributions in logit space and Dirichlet distributions (the conjugate prior to the Categorical distribution) in the output space. Importantly, the vanilla Laplace Bridge comes with certain limitations. We analyze those and suggest a simple solution that compares favorably to other commonly used estimates of the softmax-Gaussian integral. We demonstrate that the resulting Dirichlet distribution has multiple advantages, in particular, more efficient computation of the uncertainty estimate and scaling to large datasets and networks like ImageNet and DenseNet. We further demonstrate the usefulness of this Dirichlet approximation by using it to construct a lightweight uncertainty-aware output ranking for ImageNet.

Title: Sufficiency of Markov Policies for Continuous-Time Jump Markov Decision Processes Abstract: This paper extends to Continuous-Time Jump Markov Decision Processes (CTJMDP) the classic result for Markov Decision Processes stating that, for a given initial state distribution, for every policy there is a (randomized) Markov policy, which can be defined in a natural way, such that at each time instance the marginal distributions of state-action pairs for these two policies coincide. It is shown in this paper that this equality takes place for a CTJMDP if the corresponding Markov policy defines a nonexplosive jump Markov process. If this Markov process is explosive, then at each time instance the marginal probability, that a state-action pair belongs to a measurable set of state-action pairs, is not greater for the described Markov policy than the same probability for the original policy. These results are used in this paper to prove that for expected discounted total costs and for average costs per unit time, for a given initial state distribution, for each policy for a CTJMDP the described a Markov policy has the same or better performance.

Title: Online joint bid/daily budget optimization of Internet advertising campaigns Abstract: Pay-per-click advertising includes various formats (e.g., search, contextual, social) with a total investment of more than 200 billion USD per year worldwide. An advertiser is given a daily budget to allocate over several campaigns, mainly distinguishing for the ad, target, or channel. Furthermore, publishers choose the ads to display and how to allocate them employing auctioning mechanisms, in which, every day and for each campaign, the advertisers set a bid corresponding to the maximum amount of money per click they are willing to pay and the fraction of the daily budget to invest. In this paper, we study the problem of automating the online joint bid/daily budget optimization of pay-per-click advertising campaigns over multiple channels, and we face the challenging goal of designing techniques with theoretical guarantees that can be applied in real world applications, where, commonly, data scarcity is a crucial issue. We formulate our problem as a combinatorial semi-bandit problem, which requires solving a special case of the Multiple-Choice Knapsack problem every day. Furthermore, we address data scarcity by designing a model for the dependency of the number of clicks on the bid and daily budget, requiring few parameters at the cost of mild regularity assumptions. We propose two algorithms & mdash;the first is randomized, while the second is deterministic & mdash;and show that they suffer from a regret that is upper bounded with high probability as T is the time horizon of the learning process. We experimentally evaluate our algorithms with synthetic settings generated from real data provided by Yahoo!, and we present the results of adopting our algorithms in a real-world application with a daily spent of 1, 000 Euros for more than one year. O(root T), where & nbsp;T is the time horizon of the learning process. We experimentally evaluate our algorithms with synthetic settings generated from real data provided by Yahoo!, and we present the results of adopting our algorithms in a real-world application with a daily spent of 1,000 Euros for more than one year.& nbsp;(c) 2022 Elsevier B.V. All rights reserved.

Title: An efficient four-way coupled lattice Boltzmann – discrete element method for fully resolved simulations of particle-laden flows Abstract: A four-way coupling scheme for the direct numerical simulation of particle-laden flows is developed and analyzed. It employs a multiple-relaxation-time lattice Boltzmann method with an adapted bulk viscosity to simulate the fluid phase efficiently. The momentum exchange method is used to couple the fluid and the particulate phase. The particle interactions in normal and tangential direction are accounted for by a discrete element method using linear contact forces. All parameters of the scheme are studied and evaluated in detail and precise guidelines for their choice are developed. The development is based on several carefully selected calibration and validation tests of increasing physical complexity. It is found that a well-calibrated lubrication model is crucial to obtain the correct trajectories of a sphere colliding with a plane wall in a viscous fluid. For adequately resolving the collision dynamics it is found that the collision time must be stretched appropriately. The complete set of tests establishes a validation pipeline that can be universally applied to other fluid-particle coupling schemes providing a systematic methodology that can guide future developments.

Title: Admissible Orders on Fuzzy Numbers Abstract: From the more than two hundred partial orders for fuzzy numbers proposed in the literature, only a few are total. In this article, we introduce the notion of admissible order for fuzzy numbers equipped with a partial order, i.e., a total order which refines the partial order. In particular, it is given special attention to the partial order proposed by Klir and Yuan in 1995. Moreover, we propose a method to construct admissible orders on fuzzy numbers in terms of linear orders defined for intervals considering a strictly increasing upper dense sequence, proving that this order is admissible for a given partial order. Finally, we use admissible orders to ranking the path costs in fuzzy weighted graphs.

Title: Isometries between completely regular vector-valued function spaces Abstract: In this paper, first we study surjective isometries (not necessarily linear) between completely regular subspaces A and B of $$C_0(X,E)$$ and $$C_0(Y,F)$$ where X and Y are locally compact Hausdorff spaces and E and F are normed spaces, not assumed to be either strictly convex or complete. We show that for a class of normed spaces F satisfying a newly defined property related to their T-sets, such an isometry is a (generalized) weighted composition operator up to a translation. Then we apply the result to study surjective isometries between A and B whenever A and B are equipped with certain norms rather than the supremum norm. Our results unify and generalize some recent results in this context.

Title: A Quadratic Identity In The Shuffle Algebra And An Alternative Proof For De Bruijn'S Formula Abstract: Motivated by a polynomial identity of certain iterated integrals, first observed in Colmenarejo et al. (2020) in the setting of lattice paths, we prove an intriguing combinatorial identity in the shuffle algebra. It has a close connection to de Bruijn's formula when interpreted in the framework of signatures of paths. (C) 2021 Elsevier Ltd. All rights reserved.

Title: An average degree condition for independent transversals Abstract: In 1994, Erdős, Gyárfás and Łuczak posed the following problem: given disjoint vertex sets V1,…,Vn of size k, with exactly one edge between any pair Vi,Vj, how large can n be such that there will always be an independent transversal? They showed that the maximal n is at most (1+o(1))k2, by providing an explicit construction with these parameters and no independent transversal. They also proved a lower bound which is smaller by a 2e-factor.

Title: EXPLICIT BOUND FOR THE NUMBER OF PRIMES IN ARITHMETIC PROGRESSIONS ASSUMING THE GENERALIZED RIEMANN HYPOTHESIS Abstract: We prove an explicit error term for the psi(x, chi) function assuming the Generalized Riemann Hypothesis. Using this estimate, we prove a conditional explicit bound for the number of primes in arithmetic progressions.

Title: On the metric dimensions for sets of vertices. Abstract: Resolving sets were originally designed to locate vertices of a graph one at a time. For the purpose of locating multiple vertices of the graph simultaneously, $\{\ell\}$-resolving sets were recently introduced. In this paper, we present new results regarding the $\{\ell\}$-resolving sets of a graph. In addition to proving general results, we consider $\{2\}$-resolving sets in rook's graphs and connect them to block designs. We also introduce the concept of $\ell$-solid-resolving sets, which is a natural generalisation of solid-resolving sets. We prove some general bounds and characterisations for $\ell$-solid-resolving sets and show how $\ell$-solid- and $\{\ell\}$-resolving sets are connected to each other. In the last part of the paper, we focus on the infinite graph family of flower snarks. We consider the $\ell$-solid- and $\{\ell\}$-metric dimensions of flower snarks. In two proofs regarding flower snarks, we use a new computer-aided reduction-like approach.

Title: Arc-disjoint in- and out-branchings in digraphs of independence number at most 2 Abstract: We prove that every digraph of independence number at most 2 and arc-connectivity at least 2 has an out-branching B + and an in-branching B - which are arc-disjoint (we call such branchings a good pair). This is best possible in terms of the arc-connectivity as there are infinitely many strong digraphs with independence number 2 and arbitrarily high minimum in- and out-degrees that have no good pair. The result settles a conjecture by Thomassen for digraphs of independence number 2. We prove that every digraph on at most 6 vertices and arc-connectivity at least 2 has a good pair and give an example of a 2-arc-strong digraph D on 10 vertices with independence number 4 that has no good pair. We also show that there are infinitely many digraphs with independence number 7 and arc-connectivity 2 that have no good pair. Finally we pose a number of open problems.

Title: Polarization Problem on a Higher-Dimensional Sphere for a Simplex Abstract: We study the problem of maximizing the minimal value over the sphere Sd-1 subset of R-d of the potential generated by a configuration of d+1 points on Sd-1 (the maximal discrete polarization problem). The points interact via the potential given by a function f of the Euclidean distance squared, where f : [0, 4] -> (-infinity, infinity] is continuous (in the extended sense), decreasing on [0, 4], and finite and convex on (0, 4], with a concave or convex derivative f'. We prove that the configuration of the vertices of a regular d-simplex inscribed in Sd-1 is optimal. This result is new for d > 3 (certain special cases for d = 2 and d = 3 are also new). As a byproduct, we find a simpler proof for the known optimal covering property of the vertices of a regular d-simplex inscribed in Sd-1.

Title: Scheduling Kernels via Configuration LP Abstract: Makespan minimization (on parallel identical or unrelated machines) is arguably the most natural and studied scheduling problem. A common approach in practical algorithm design is to reduce the size of a given instance by a fast preprocessing step while being able to recover key information even after this reduction. This notion is formally studied as kernelization (or simply, kernel) -- a polynomial time procedure which yields an equivalent instance whose size is bounded in terms of some given parameter. It follows from known results that makespan minimization parameterized by the longest job processing time $p_{\max}$ has a kernelization yielding a reduced instance whose size is exponential in $p_{\max}$. Can this be reduced to polynomial in $p_{\max}$? We answer this affirmatively not only for makespan minimization, but also for the (more complicated) objective of minimizing the weighted sum of completion times, also in the setting of unrelated machines when the number of machine kinds is a parameter. Our algorithm first solves the Configuration LP and based on its solution constructs a solution of an intermediate problem, called huge $N$-fold integer programming. This solution is further reduced in size by a series of steps, until its encoding length is polynomial in the parameters. Then, we show that huge $N$-fold IP is in NP, which implies that there is a polynomial reduction back to our scheduling problem, yielding a kernel. Our technique is highly novel in the context of kernelization, and our structural theorem about the Configuration LP is of independent interest. Moreover, we show a polynomial kernel for huge $N$-fold IP conditional on whether the so-called separation subproblem can be solved in polynomial time. Considering that integer programming does not admit polynomial kernels except for quite restricted cases, our "conditional kernel" provides new insight.

Title: Equivalence relations and determinacy Abstract: We introduce the notion of (Gamma, E)-determinacy for Gamma a pointclass and E an equivalence relation on a Polish space X. A case of particular interest is the case when E = E-G is the (left) shift-action of G on S-G where S = 2 = {0, 1} or S = omega. We show that for all shift actions by countable groups G, and any "reasonable" pointclass Gamma, that (Gamma, E-G)-determinacy implies Gamma-determinacy. We also prove a corresponding result when E is a subshift of finite type of the shift map on 2(Z).

Title: Optimally adaptive Bayesian spectral density estimation for stationary and nonstationary processes Abstract: This article improves on existing Bayesian methods to estimate the spectral density of stationary and nonstationary time series assuming a Gaussian process prior. By optimising an appropriate eigendecomposition using a smoothing spline covariance structure, our method more appropriately models data with both simple and complex periodic structure. We further justify the utility of this optimal eigendecomposition by investigating the performance of alternative covariance functions other than smoothing splines. We show that the optimal eigendecomposition provides a material improvement, while the other covariance functions under examination do not, all performing comparatively well as the smoothing spline. During our computational investigation, we introduce new validation metrics for the spectral density estimate, inspired from the physical sciences. We validate our models in an extensive simulation study and demonstrate superior performance with real data.

Title: End-to-End Trainable One-Stage Parking Slot Detection Integrating Global and Local Information Abstract: This paper proposes an end-to-end trainable one-stage parking slot detection method for around view monitor (AVM) images. The proposed method simultaneously acquires global information (entrance, type, and occupancy of parking slot) and local information (location and orientation of junction) by using a convolutional neural network (CNN), and integrates them to detect parking slots with their prop...

Title: A comparison of maximum likelihood and absolute moments for the estimation of Hurst exponents in a stationary framework Abstract: The absolute-moment method is widespread for estimating the Hurst exponent of a fractional Brownian motion X. But this method is biased when applied to a stationary version of X, in particular an inverse Lamperti transform of X, with a linear time contraction of parameter theta. We present an adaptation of the absolute-moment method to this framework and we compare it to the maximum likelihood method, with simulations and an application to a financial time series. While it appears that the maximum-likelihood method is more accurate than the adapted absolute-moment estimation, this last method is not uninteresting for two reasons: it makes it possible to confirm visually that the model is well specified and it is computationally more performing. (C) 2022 Elsevier B.V. All rights reserved.

Title: SUBCOMPACT CARDINALS, TYPE OMISSION, AND LADDER SYSTEMS Abstract: We provide a model theoretical and tree property-like characterization of lambda-pi(1)(1)- subcompactness and supercompactness. We explore the behavior of these combinatorial principles at accessible cardinals.

Title: Moduli Spaces of Codimension-One Subspaces in a Linear Variety and their Tropicalization. Abstract: We study the moduli space of $d$-dimensional linear subspaces contained in a fixed $(d+1)$-dimensional linear variety $X$, and its tropicalization. We prove that these moduli spaces are linear subspaces themselves, and thus their tropicalization is completely determined by their associated (valuated) matroids. We show that these matroids can be interpreted as the matroid of lines of the hyperplane arrangement corresponding to $X$, and generically are equal to a Dilworth truncation of the free matroid. In this way, we can describe combinatorially tropicalized Fano schemes and tropicalizations of moduli spaces of stable maps of degree $1$ to a plane.

Title: On the Ehrhart Polynomial of Minimal Matroids Abstract: We provide a formula for the Ehrhart polynomial of the connected matroid of size n and rank k with the least number of bases, also known as a minimal matroid. We prove that their polytopes are Ehrhart positive and $$h^*$$ -real-rooted (and hence unimodal). We prove that the operation of circuit-hyperplane relaxation relates minimal matroids and matroid polytopes subdivisions, and also preserves Ehrhart positivity. We state two conjectures: that indeed all matroids are $$h^*$$ -real-rooted, and that the coefficients of the Ehrhart polynomial of a connected matroid of fixed rank and cardinality are bounded by those of the corresponding minimal matroid and the corresponding uniform matroid.

Title: Real-Time Federated Evolutionary Neural Architecture Search Abstract: Federated learning is a distributed machine learning approach to privacy preservation and two major technical challenges prevent a wider application of federated learning. One is that federated learning raises high demands on communication resources, since a large number of model parameters must be transmitted between the server and clients. The other challenge is that training large machine learn...

Title: Analysis and control of a non-local PDE traffic flow model Abstract: This paper provides conditions that guarantee existence and uniqueness of classical solutions for a non-local conservation law on a ring road with possible nudging (or 'look behind') terms. The obtained conditions are novel, as they are not covered by existing results in the literature. This paper also provides results which indicate that nudging can increase the flow in a ring road and, if properly designed, can have a strong stabilising effect on traffic flow. More specifically, this paper gives results that guarantee local exponential stability of the uniform equilibrium profile in thestate norm even for cases where the uniform equilibrium profile in a ring road without nudging is not asymptotically stable and the model admits density waves. The efficiency of the use of nudging terms is demonstrated by means of a numerical example.

Title: On the border of the amyloidogenic sequences: prefix analysis of the parallel beta sheets in the PDB_Amyloid collection Abstract: The Protein Data Bank (PDB) today contains more than 174,000 entries with the 3-dimensional structures of biological macromolecules. Using the rich resources of this repository, it is possible identifying subsets with specific, interesting properties for different applications. Our research group prepared an automatically updated list of amyloid- and probably amyloidogenic molecules, the PDB Amyloid collection, which is freely available at the address http://pitgroup.orgiamyloid. This resource applies exclusively the geometric properties of the steric structures for identifying amyloids. In the present contribution, we analyze the starting (i.e., prefix) subsequences of the characteristic, parallel beta-sheets of the structures in the PDB Amyloid collection, and identify further appearances of these length-5 prefix subsequences in the whole PDB data set. We have identified this way numerous proteins, whose normal or irregular functions involve amyloid formation, structural misfolding, or anti-coagulant properties, simply by containing these prefixes: including the T-cell receptor (TCR), bound with the major histocompatibility complexes MHC-1 and MHC-2; the p53 tumor suppressor protein; a mycobacterial RNA polymerase transcription initialization complex; the human bridging integrator protein BIN-1; and the tick anti-coagulant peptide TAP.

Title: A MEAN FIELD GAME APPROACH TO EQUILIBRIUM PRICING WITH MARKET CLEARING CONDITION Abstract: In this work, we study an equilibrium-based continuous asset pricing problem which seeks to form a price process endogenously by requiring it to balance the flow of sale-and-purchase orders in the exchange market, where a large number of agents 1 \leq i \leq N are interacting through the market price. Adopting a mean field game approach, we find a special form of forward-backward stochastic differential equations of McKean-Vlasov type with common noise whose solution provides an approximate of the market price. We show the convergence of the net order flow to zero in the large N-limit and get the order of convergence in N under some conditions.

Title: Cost-Sensitive Portfolio Selection via Deep Reinforcement Learning Abstract: Portfolio Selection is an important real-world financial task and has attracted extensive attention in artificial intelligence communities. This task, however, has two main difficulties: (i) the non-stationary price series and complex asset correlations make the learning of feature representation very hard; (ii) the practicality principle in financial markets requires controlling both transaction ...

Title: GeoConv: Geodesic guided convolution for facial action unit recognition Abstract: •We propose a novel geodesic guided convolution (GeoConv), which can be applied to fine-grained face analysis tasks such as AU recognition.•We propose an AU recognition network based on GeoConv. This is the first work of integrating 3D information into 2D convolution for AU recognition.•Experiments verify the effectiveness of our approach and show that our method soundly outperforms the SOTA methods on BP4D and DISFA benchmarks.

Title: Expected complexity analysis of stochastic direct-search Abstract: This work presents the convergence rate analysis of stochastic variants of the broad class of direct-search methods of directional type. It introduces an algorithm designed to optimize differentiable objective functions f whose values can only be computed through a stochastically noisy blackbox. The proposed stochastic directional direct-search (SDDS) algorithm accepts new iterates by imposing a sufficient decrease condition on so called probabilistic estimates of the corresponding unavailable objective function values. The accuracy of such estimates is required to hold with a sufficiently large but fixed probability beta. The analysis of this method utilizes an existing supermartingale-based framework proposed for the convergence rates analysis of stochastic optimization methods that use adaptive step sizes. It aims to show that the expected number of iterations required to drive the norm of the gradient of f below a given threshold epsilon is bounded in O((epsilon) -p/min(p-1,1) /(2 beta - 1)) with p > 1. Unlike prior analysis using the same aforementioned framework such as those of stochastic trust-region methods and stochastic line search methods, SDDS does not use any gradient information to find descent directions. However, its convergence rate is similar to those of both latter methods with a dependence on epsilon that also matches that of the broad class of deterministic directional direct-search methods which accept new iterates by imposing a sufficient decrease condition.

Title: Automated repair of resource leaks in Android applications Abstract: Resource leaks – a program does not release resources it previously acquired – are a common kind of bug in Android applications. Even with the help of existing techniques to automatically detect leaks, writing a leak-free program remains tricky. One of the reasons is Android’s event-driven programming model, which complicates the understanding of an application’s overall control flow. In this paper, we present : a technique to automatically detect and fix resource leaks in Android applications.  builds a succinct abstraction of an app’s control flow, and uses it to find execution traces that may leak a resource. The information built during detection also enables automatically building a fix – consisting of release operations performed at appropriate locations – that removes the leak and does not otherwise affect the application’s usage of the resource. An empirical evaluation on resource leaks from the curated collection demonstrates that ’s approach is scalable, precise, and produces correct fixes for a variety of resource leak bugs:  automatically found and repaired 50 leaks that affect 9 widely used resources of the Android system, including all those collected by for those resources; on average, it took just 2 min to detect and repair a leak. also compares favorably to Relda2/RelFix – the only other fully automated approach to repair Android resource leaks – since it can often detect more leaks with higher precision and producing smaller fixes. These results indicate that can provide valuable support to enhance the quality of Android applications in practice.

Title: Keyless Covert Communication via Channel State Information Abstract: We consider the problem of covert communication over a state-dependent channel when the Channel State Information (CSI) is available either non-causally, causally, or strictly causally, either at the transmitter alone, or at both transmitter and receiver. Covert communication with respect to an adversary, called “warden,” is one in which, despite communication over the channel, the warden’s observation remains indistinguishable from an output induced by innocent channel-input symbols. Covert communication involves fooling an adversary in part by a proliferation of codebooks; for reliable decoding at the legitimate receiver, the codebook uncertainty is typically removed via a shared secret key that is unavailable to the warden. In contrast to previous work, we do not assume the availability of a large shared key at the transmitter and legitimate receiver. Instead, we only require a secret key with negligible rate to bootstrap the communication and our scheme extracts shared randomness from the CSI in a manner that keeps it secret from the warden, despite the influence of the CSI on the warden’s output. When CSI is available at the transmitter and receiver, we derive the covert capacity region. When CSI is only available at the transmitter, we derive inner and outer bounds on the covert capacity. We also provide examples for which the covert capacity is positive with knowledge of CSI but is zero without it.

Title: Covering cycles in sparse graphs Abstract: Let k >= 2 be an integer. Kouider and Lonc proved that the vertex set of every graph G with n >= n0(k) vertices and minimum degree at least n/k can be covered by k-1 cycles. Our main result states that for every alpha>0 and p=p(n)is an element of(0,1], the same conclusion holds for graphs G with minimum degree (1/k+alpha)np that are sparse in the sense that eG(X,Y)<= p|X||Y|+o(np|X||Y|/log3n)for all X,Y subset of V(G). In particular, this allows us to determine the local resilience of random and pseudorandom graphs with respect to having a vertex cover by a fixed number of cycles. The proof uses a version of the absorbing method in sparse expander graphs.

Title: Smart Train Operation Algorithms Based on Expert Knowledge and Reinforcement Learning Abstract: During decades, the automatic train operation (ATO) system has been gradually adopted in many subway systems for its low-cost and intelligence. This article proposes two smart train operation (STO) algorithms by integrating the expert knowledge with reinforcement learning algorithms. Compared with previous works, the proposed algorithms can realize the control of continuous action for the subway system and optimize multiple critical objectives without using an offline speed profile. First, through learning historical data of experienced subway drivers, we extract the expert knowledge rules and build inference methods to guarantee the riding comfort, the punctuality, and the safety of the subway system. Then we develop two algorithms for optimizing the energy efficiency of train operation. One is the STO algorithm based on deep deterministic policy gradient named (STOD) and the other is the STO algorithm based on normalized advantage function (STON). Finally, we verify the performance of proposed algorithms via some numerical simulations with the real field data from the Yizhuang Line of the Beijing Subway and illustrate that the developed STO algorithm are better than expert manual driving and existing ATO algorithms in terms of energy efficiency. Moreover, STOD and STON can adapt to different trip times and different resistance conditions.

Title: New Constructions of Complementary Sequence Pairs Over 4 q -QAM Abstract: The researches of Golay complementary sequences (GCSs) over 16 and 64 quadrature amplitude modulation (QAM) from 2001 to 2008 were generalized to $4^{q} $ -QAM GCSs of length $2^{m}$ by Li (the generalized cases I-III for <tex-math notati...

Title: Deep learning for prediction of population health costs Abstract: Background Accurate prediction of healthcare costs is important for optimally managing health costs. However, methods leveraging the medical richness from data such as health insurance claims or electronic health records are missing. Methods Here, we developed a deep neural network to predict future cost from health insurance claims records. We applied the deep network and a ridge regression model to a sample of 1.4 million German insurants to predict total one-year health care costs. Both methods were compared to existing models with various performance measures and were also used to predict patients with a change in costs and to identify relevant codes for this prediction. Results We showed that the neural network outperformed the ridge regression as well as all considered models for cost prediction. Further, the neural network was superior to ridge regression in predicting patients with cost change and identified more specific codes. Conclusion In summary, we showed that our deep neural network can leverage the full complexity of the patient records and outperforms standard approaches. We suggest that the better performance is due to the ability to incorporate complex interactions in the model and that the model might also be used for predicting other health phenotypes.

Title: Learn and Transfer Knowledge of Preferred Assistance Strategies in Semi-Autonomous Telemanipulation Abstract: Enabling robots to provide effective assistance yet still accommodating the operator’s commands for telemanipulation of an object is very challenging because robot’s assistance is not always intuitive for human operators and human behaviors and preferences are sometimes ambiguous for the robot to interpret. Due to the difference in hand structures, some motion assistance from the robot may surprise the operator with counter-intuitive movements, which could introduce more burden to the human to correct the actions and/or reduce the operator’s sense of system control. To address these problems, we developed a novel preference-aware assistance knowledge learning approach. An assistance preference model learns what assistance is preferred by a human, and a stage-wise model updating method ensures the learning stability while dealing with the ambiguity of human preference data. Such a preference-aware assistance knowledge enables a teleoperated robot hand to provide more active yet preferred assistance toward manipulation success. We also developed knowledge transfer methods to transfer the preference knowledge across different robot hand structures to avoid extensive robot-specific training. Experiments to telemanipulate a 3-finger hand and 2-finger hand, respectively, to use, move, and hand over a cup have been conducted. Results demonstrated that the methods enabled the robots to effectively learn the preference knowledge and allowed knowledge transfer between robots with less training effort.

Title: Ternary Compression for Communication-Efficient Federated Learning Abstract: Learning over massive data stored in different locations is essential in many real-world applications. However, sharing data is full of challenges due to the increasing demands of privacy and security with the growing use of smart mobile devices and Internet of thing (IoT) devices. Federated learning provides a potential solution to privacy-preserving and secure machine learning, by means of jointly training a global model without uploading data distributed on multiple devices to a central server. However, most existing work on federated learning adopts machine learning models with full-precision weights, and almost all these models contain a large number of redundant parameters that do not need to be transmitted to the server, consuming an excessive amount of communication costs. To address this issue, we propose a federated trained ternary quantization (FTTQ) algorithm, which optimizes the quantized networks on the clients through a self-learning quantization factor. Theoretical proofs of the convergence of quantization factors, unbiasedness of FTTQ, as well as a reduced weight divergence are given. On the basis of FTTQ, we propose a ternary federated averaging protocol (T-FedAvg) to reduce the upstream and downstream communication of federated learning systems. Empirical experiments are conducted to train widely used deep learning models on publicly available data sets, and our results demonstrate that the proposed T-FedAvg is effective in reducing communication costs and can even achieve slightly better performance on non-IID data in contrast to the canonical federated learning algorithms.

Title: Securing LSB embedding against structural steganalysis Abstract: This work explores the extent to which LSB embedding can be made secure against structural steganalysis through a modification of cover image statistics prior to message embedding. LSB embedding disturbs the statistics of consecutive k-tuples of pixels, and a kth-order structural attack detects hidden messages with lengths in proportion to the size of the imbalance amongst sets of k-tuples. To protect against kth-order structural attacks, cover modifications involve the redistribution of k-tuples among the different sets so that symmetries of the cover image are broken, then repaired through the act of LSB embedding so that the stego image bears the statistics of the original cover. We find this is only feasible for securing against up to 3rd-order attacks since higher-order protections result in virtually zero embedding capacities. To protect against 3rd-order attacks, we perform a redistribution of triplets that also preserves the statistics of pairs. This is done by embedding into only certain pixels of each sextuplet, constraining the maximum embedding rate to be <= 2/3 bits per channel. Testing on a variety of image formats, we report best performance for JPEG-compressed images with a mean maximum embedding rate undetectable by 2nd- and 3rd-order attacks of 0.21 bpc.

Title: Dynamic Backdoor Attacks Against Machine Learning Models Abstract: Machine learning (ML) has made tremendous progress during the past decade and is being adopted in various critical real-world applications. However, recent research has shown that ML models are vulnerable to multiple security and privacy attacks. In particular, backdoor attacks against ML models have recently raised a lot of awareness. A successful backdoor attack can cause severe consequences, such as allowing an adversary to bypass critical authentication systems. Current backdooring techniques rely on adding static triggers (with fixed patterns and locations) on ML model inputs which are prone to detection by the current backdoor detection mechanisms. In this paper, we propose the first class of dynamic backdooring techniques against deep neural networks (DNN), namely Random Backdoor, Backdoor Generating Network (BaN), and conditional Backdoor Generating Network (c-BaN). Triggers generated by our techniques can have random patterns and locations, which reduce the efficacy of the current backdoor detection mechanisms. In particular, BaN and c-BaN based on a novel generative network are the first two schemes that algorithmically generate triggers. Moreover, c-BaN is the first conditional backdooring technique that given a target label, it can generate a target-specific trigger. Both BaN and c-BaN are essentially a general framework which renders the adversary the flexibility for further customizing backdoor attacks. We extensively evaluate our techniques on three benchmark datasets: MNIST, CelebA, and CIFAR-10. Our techniques achieve almost perfect attack performance on back-doored data with a negligible utility loss. We further show that our techniques can bypass current state-of-the-art defense mechanisms against backdoor attacks, including ABS, Februus, MNTD, Neural Cleanse, and STRIP.

Title: Min–Max Q-learning for multi-player pursuit-evasion games Abstract: In this paper, we address a pursuit-evasion game involving multiple players by utilizing tools and techniques from reinforcement learning and matrix game theory. In particular, we consider the problem of steering an evader to a goal destination while avoiding capture by multiple pursuers, which is a high-dimensional and computationally intractable problem in general. In our proposed approach, we first formulate the multi-agent pursuit-evasion game as a sequence of discrete matrix games. Next, in order to simplify the solution process, we transform the high-dimensional state space into a low-dimensional manifold and the continuous action space into a feature-based space, which is a discrete abstraction of the original space. Based on these transformed state and action spaces, we subsequently employ min–max Q-learning, to generate the entries of the payoff matrix of the game, and subsequently obtain the optimal action for the evader at each stage. Finally, we present extensive numerical simulations to evaluate the performance of the proposed learning-based evading strategy in terms of the evader’s ability to reach the desired target location without being captured, as well as computational efficiency.

Title: Active Fine-Tuning From gMAD Examples Improves Blind Image Quality Assessment Abstract: The research in image quality assessment (IQA) has a long history, and significant progress has been made by leveraging recent advances in deep neural networks (DNNs). Despite high correlation numbers on existing IQA datasets, DNN-based models may be easily falsified in the group maximum differentiation (gMAD) competition. Here we show that gMAD examples can be used to improve blind IQA (BIQA) methods. Specifically, we first pre-train a DNN-based BIQA model using multiple noisy annotators, and fine-tune it on multiple synthetically distorted images, resulting in a “top-performing” baseline model. We then seek pairs of images by comparing the baseline model with a set of full-reference IQA methods in gMAD. The spotted gMAD examples are most likely to reveal the weaknesses of the baseline, and suggest potential ways for refinement. We query human quality annotations for the selected images in a well-controlled laboratory environment, and further fine-tune the baseline on the combination of human-rated images from gMAD and existing databases. This process may be iterated, enabling active fine-tuning from gMAD examples for BIQA. We demonstrate the feasibility of our active learning scheme on a large-scale unlabeled image set, and show that the fine-tuned quality model achieves improved generalizability in gMAD, without destroying performance on previously seen databases.

Title: Online data-enabled predictive controlx2729; Abstract: We develop an online data-enabled predictive (ODeePC) control method for optimal control of unknown systems, building on the recently proposed DeePC (Coulson et al., 2019). Our proposed ODeePC method leverages a primal-dual algorithm with real-time measurement feedback to iteratively compute the corresponding real-time optimal control policy as system conditions change. The proposed ODeePC conceptual-wise resembles standard adaptive system identification and model predictive control (MPC), but it provides a new alternative for the standard methods. ODeePC is enabled by computationally efficient methods that exploit the special structure of the Hankel matrices in the context of DeePC with Fast Fourier Transform (FFT) and primal-dual algorithm We provide theoretical guarantees regarding the asymptotic behavior of ODeePC, and we demonstrate its performance through numerical examples.(C) 2021 Published by Elsevier Ltd.

Title: Divided Differences, Falling Factorials, and Discrete Splines: Another Look at Trend Filtering and Related Problems. Abstract: This paper serves as a postscript of sorts to Tibshirani (2014); Wang et al. (2014), who developed continuous-time formulations and properties of trend filtering, a discrete-time smoothing tool proposed (independently) by Steidl et al. (2006); Kim et al. (2009). The central object of study is the falling factorial basis, as it was called by Tibshirani (2014); Wang et al. (2014). Its span turns out to be a space of piecewise polynomials that has a classical place in spline theory, called discrete splines (Mangasarian and Schumaker, 1971, 1973; Schumaker, 2007). At the Tibshirani (2014); Wang et al. (2014), we were not fully aware of these connections. The current paper attempts to rectify this by making these connections explicit, reviewing (and making use of) some of the important existing work on discrete splines, and contributing several new perspectives and new results on discrete splines along the way.

Title: ROSE: real one-stage effort to detect the fingerprint singular point based on multi-scale spatial attention Abstract: Detecting the singular point accurately and efficiently is one of the most important tasks for fingerprint recognition. In recent years, deep learning has been gradually used in the fingerprint singular point detection. However, the existing deep learning-based singular point detection methods are either two-stage or multi-stage, which makes them time-consuming. More importantly, their detection accuracy is yet unsatisfactory, especially for the low-quality fingerprint. In this paper, we make a Real One-Stage Effort to detect fingerprint singular points more accurately and efficiently, and therefore, we name the proposed algorithm ROSE for short, in which the multi-scale spatial attention, the Gaussian heatmap and the variant of focal loss are integrated together to achieve a higher detection rate. Experimental results on the datasets FVC2002 DB1 and NIST SD4 show that our ROSE outperforms the state-of-the-art algorithms in terms of detection rate, false alarm rate and detection speed.

Title: Cross-Modal Food Retrieval: Learning a Joint Embedding of Food Images and Recipes With Semantic Consistency and Attention Mechanism Abstract: Food retrieval is an important task to perform analysis of food-related information, where we are interested in retrieving relevant information about the queried food item such as ingredients, cooking instructions, etc. In this paper, we investigate cross-modal retrieval between food images and cooking recipes. The goal is to learn an embedding of images and recipes in a common feature space, such that the corresponding image-recipe embeddings lie close to one another. Two major challenges in addressing this problem are 1) large intra-variance and small inter-variance across cross-modal food data; and 2) difficulties in obtaining discriminative recipe representations. To address these two problems, we propose Semantic-Consistent and Attention-based Networks (SCAN), which regularize the embeddings of the two modalities through aligning output semantic probabilities. Besides, we exploit a self-attention mechanism to improve the embedding of recipes. We evaluate the performance of the proposed method on the large-scale Recipe1M dataset, and show that we can outperform several state-of-the-art cross-modal retrieval strategies for food images and cooking recipes by a significant margin.

Title: evgam: An R Package for Generalized Additive Extreme Value Models Abstract: This article introduces the R package evgam. The package provides functions for fitting extreme value distributions. These include the generalized extreme value and generalized Pareto distributions. The former can also be fitted through a point process representation. Package evgam supports quantile regression via the asymmetric Laplace distribution, which can be useful for estimating high thresholds, sometimes used to discriminate between extreme and non-extreme values. The main addition of package evgam is to let extreme value distribution parameters have generalized additive model forms, the smoothness of which can be objectively estimated using Laplace's method. Illustrative examples fitting various distributions with various specifications are given. These include daily precipitation accumulations for part of Colorado, US, used to illustrate spatial models, and daily maximum temperatures for Fort Collins, Colorado, US, used to illustrate temporal models.

Title: A multi-source entity-level sentiment corpus for the financial domain: the FinLin corpus Abstract: We introduce FinLin, a novel corpus containing investor reports, company reports, news articles, and microblogs from StockTwits, targeting multiple entities stemming from the automobile industry and covering a 3-month period. FinLin was annotated with a sentiment score and a relevance score in the range [− 1.0, 1.0] and [0.0, 1.0], respectively. The annotations also include the text spans selected for the sentiment, thus, providing additional insight into the annotators’ reasoning. Overall, FinLin aims to complement the current knowledge by providing a novel and publicly available financial sentiment corpus and to foster research on the topic of financial sentiment analysis and potential applications in behavioural science.

Title: Change Point Models for Real-Time Cyber Attack Detection in Connected Vehicle Environment Abstract: Connected vehicle (CV) systems are subject to potential cyber attacks because of increasing connectivity between its different components, such as vehicles, roadside infrastructure, and traffic management centers. However, it is a challenge to detect security threats in real-time and develop appropriate or effective countermeasures for a CV system because of the dynamic behavior of such attacks, high computational power requirement, and a historical data requirement for training detection models. To address these challenges, statistical models, especially change point models, have potentials for real-time anomaly detection. Thus, the objective of this study is to investigate the efficacy of two change point models; Expectation Maximization (EM) and two forms of Cumulative Summation (CUSUM) algorithms (i.e., typical and adaptive), for real-time vehicle-to-infrastructure (V2I) cyber attack detection in a CV Environment. To prove the efficacy of these models, we evaluated these two models for three different types of cyber attack, denial of service (DOS), impersonation, and false information, using basic safety messages (BSMs) generated from CVs through simulation. Results from numerical analysis revealed that EM, CUSUM, and adaptive CUSUM (aCUSUM) could detect these cyberattacks, such as DOS, impersonation, and false information with low false positives.

Title: Deep neural networks for automatic speech processing: a survey from large corpora to limited data Abstract: Most state-of-the-art speech systems use deep neural networks (DNNs). These systems require a large amount of data to be learned. Hence, training state-of-the-art frameworks on under-resourced speech challenges are difficult tasks. As an example, a challenge could be the limited amount of data to model impaired speech. Furthermore, acquiring more data and/or expertise is time-consuming and expensive. In this paper, we focus on the following speech processing tasks: automatic speech recognition, speaker identification, and emotion recognition. To assess the problem of limited data, we firstly investigate state-of-the-art automatic speech recognition systems, as this is the hardest task (due to the wide variability in each language). Next, we provide an overview of techniques and tasks requiring fewer data. In the last section, we investigate few-shot techniques by interpreting under-resourced speech as a few-shot problem. In that sense, we propose an overview of few-shot techniques and the possibility of using such techniques for the speech problems addressed in this survey. It is true that the reviewed techniques are not well adapted for large datasets. Nevertheless, some promising results from the literature encourage the usage of such techniques for speech processing.

Title: FusionLane: Multi-Sensor Fusion for Lane Marking Semantic Segmentation Using Deep Neural Networks Abstract: Effective semantic segmentation of lane marking is crucial for construction of high-precision lane level maps. In recent years, a number of different methods for semantic segmentation of images have been proposed. These methods concentrate mainly on analysis of camera images, due to limitations with the sensor itself, and thus far, the accurate three-dimensional spatial position of the lane marking could not be obtained, which hinders lane level map construction.This article proposes a lane marking semantic segmentation method based on LIDAR and camera image fusion using a deep neural network. In the approach, the object of the semantic segmentation is a bird’s-eye view converted from a LIDAR points cloud instead of an image captured by a camera. First, the DeepLabV3+ network image segmentation method is used to segment the image captured by the camera, and the segmentation result is then merged with the point clouds collected by the LIDAR as the input of the proposed network. A long short-term memory (LSTM) structure is added to the neural network to assist the network in semantic segmentation of lane markings by enabling use of time series information. Experiments on datasets containing more than 14,000 images, which were manually labeled and expanded, showed that the proposed method provides accurate semantic segmentation of the bird’s-eye view LIDAR points cloud. Consequently, automation of high-precision map construction can be significantly improved. Our code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/rolandying/FusionLane</uri> .

Title: Correlated Initialization for Correlated Data Abstract: Spatial data exhibits the property that nearby points are correlated. This also holds for learnt representations across layers, but not for commonly used weight initialization methods. Our theoretical analysis quantifies the learning behavior of weights of a single spatial filter. It is thus in contrast to a large body of work that discusses statistical properties of weights. It shows that uncorrelated initialization (1) might lead to poor convergence behavior and (2) training of (some) parameters is likely subject to slow convergence. Empirical analysis shows that these findings for a single spatial filter extend to networks with many spatial filters. The impact of (correlated) initialization depends strongly on learning rates and l2-regularization.

Title: A Mixed Finite Element Discretization of Dynamical Optimal Transport Abstract: In this paper we introduce a new class of finite element discretizations of the quadratic optimal transport problem based on its dynamical formulation. These generalize to the finite element setting the finite difference scheme proposed by Papadakis et al. (SIAM J Imaging Sci, 7(1):212–238, 2014). We solve the discrete problem using a proximal splitting approach and we show how to modify this in the presence of regularization terms which are relevant for physical data interpolation.

Title: The Locus Algorithm: A novel technique for identifying optimised pointings for differential photometry Abstract: Studies of the photometric variability of astronomical sources from ground-based telescopes must overcome atmospheric extinction effects. Differential photometry by reference to an ensemble of reference stars which closely match the target in terms of magnitude and colour can mitigate these effects. This Paper describes the design, implementation, and operation of a novel algorithm - The Locus Algorithm - which enables optimised differential photometry. The Algorithm is intended to identify, for a given target and observational parameters, the Field of View (FoV) which includes the target and the maximum number of reference stars similar to the target. A collection of objects from a catalogue (e.g. SDSS) is filtered to identify candidate reference stars and determine a rating for each which quantifies its similarity to the target. The algorithm works by defining a locus of points around each candidate reference star, upon which the FoV can be centred and include the reference at the edge of the FoV. The Points of Intersection (PoI) between these loci are identified and a score for each PoI is calculated. The PoI with the highest score is output as the optimum pointing. The steps of the algorithm are precisely defined in this paper. The application of The Locus Algorithm to a sample target, SDSS1237680117417115655, from the Sloan Digital Sky Survey is described in detail. The algorithm has been defined here and implemented in software which is available online. The algorithm has also been used to generate catalogues of pointings to optimise Quasar variability studies and to generate catalogues of optimised pointings in the search for Exoplanets via the transit method.(C) 2021 Elsevier B.V. All rights reserved.

Title: Adjunction in the Absence of Identity Abstract: We develop a bicategorical setup in which one can speak about adjoint 1-morphisms even in the absence of genuine identity 1-morphisms. We also investigate which part of 2-representation theory of 2-categories extends to this new setup.

Title: Point Partition Numbers: Perfect Graphs Abstract: Graphs considered in this paper are finite, undirected and without loops, but with multiple edges. For an integer t >= 1, denote by Mg-t the class of graphs whose maximum multiplicity is at most t. A graph G is called strictly t-degenerate if every non-empty subgraph H of G contains a vertex v whose degree in H is at most t - 1. The point partition number chi(t)(G) of G is the smallest number of colors needed to color the vertices of G so that each vertex receives a color and vertices with the same color induce a strictly t-degenerate subgraph of G. So chi(1) is the chromatic number, and chi(2) is known as the point aboricity. The point partition number chi(t) with t >= 1 was introduced by Lick and White (Can J Math 22:1082-1096, 1970). If H is a simple graph, then tH denotes the graph obtained from H by replacing each edge of H by t parallel edges. Then omega(t)(G) is the largest integer n such that G contains a tK(n) as a subgraph. Let G be a graph belonging to Mg-t. Then omega(t)(G) <= chi(t)(G) and we say that G is chi(t)-perfect if every induced subgraph H of G satisfies omega(t)(H) = chi(t)(H). Based on the Strong Perfect Graph Theorem due to Chudnowsky, Robertson, Seymour and Thomas (Ann Math 164:51-229, 2006), we give a characterization of chi(t)-perfect graphs of Mg-t by a set of forbidden induced subgraphs (see Theorems 2 and 3). We also discuss some complexity problems for the class of chi(t)-critical graphs.

Title: Optimizing revenue while showing relevant assortments at scale Abstract: •Solves the assortment optimization problem under the multinomial logit model.•Proposes scalable algorithms that choose assortments from an arbitrary collection.•Establishes connection between the combinatorial problem and nearest neighbor search.•Provides time-complexity guarantees for general and capacity constrained settings.

Title: Multivariate Functional Regression Via Nested Reduced-Rank Regularization Abstract: We propose a nested reduced-rank regression (NRRR) approach in fitting a regression model with multivariate functional responses and predictors to achieve tailored dimension reduction and facilitate model interpretation and visualization. Our approach is based on a two-level low-rank structure imposed on the functional regression surfaces. A global low-rank structure identifies a small set of latent principal functional responses and predictors that drives the underlying regression association. A local low-rank structure then controls the complexity and smoothness of the association between the principal functional responses and predictors. The functional problem boils down to an integrated matrix approximation task through basis expansion, where the blocks of an integrated low-rank matrix share some common row space and/or column space. This nested reduced-rank structure also finds potential applications in multivariate time series modeling and tensor regression. A blockwise coordinate descent algorithm is developed. We establish the consistency of NRRR and show through nonasymptotic analysis that it can achieve at least a comparable error rate to that of the reduced-rank regression. Simulation studies demonstrate the effectiveness of NRRR. We apply the proposed methods in an electricity demand problem to relate daily electricity consumption trajectories with daily temperatures. Supplementary files for this article are available online.

Title: HEMlets PoSh: Learning Part-Centric Heatmap Triplets for 3D Human Pose and Shape Estimation Abstract: Estimating 3D human pose from a single image is a challenging task. This work attempts to address the uncertainty of lifting the detected 2D joints to the 3D space by introducing an intermediate state - Part-Centric Heatmap Triplets (HEMlets), which shortens the gap between the 2D observation and the 3D interpretation. The HEMlets utilize three joint-heatmaps to represent the relative depth...

Title: Kernel Quantization for Efficient Network Compression Abstract: This paper presents a novel network compression framework, Kernel Quantization (KQ), targeting to efficiently convert any pre-trained full-precision convolutional neural network (CNN) model into a low-precision version without significant performance loss. Unlike existing methods struggling with weight bit-length, KQ has the potential in improving the compression ratio by considering the convolution kernel as the quantization unit. Inspired by the evolution from weight pruning to filter pruning, we propose to quantize in both kernel and weight level. Instead of representing each weight parameter with a low-bit index, we learn a kernel codebook and replace all kernels in the convolution layer with corresponding low-bit indexes. Thus, KQ can represent the weight tensor in the convolution layer with low-bit indexes and a kernel codebook with limited size, which enables KQ to achieve significant compression ratio. Then, we conduct a 6-bit parameter quantization on the kernel codebook to further reduce redundancy. Extensive experiments on the ImageNet classification task prove that KQ needs 1.05 and 1.62 bits on average in VGG and ResNet18, respectively, to represent each parameter in the convolution layer and achieves the state-of-the-art compression ratio with little accuracy loss.

Title: Colorings, transversals, and local sparsity Abstract: Motivated both by recently introduced forms of list coloring and by earlier work on independent transversals subject to a local sparsity condition, we use the semi-random method to prove the following result. For any function mu satisfying mu(d)=o(d) as d ->infinity, there is a function lambda satisfying lambda(d)=d+o(d) as d ->infinity such that the following holds. For any graph H and any partition of its vertices into parts of size at least lambda such that (a) for each part the average over its vertices of degree to other parts is at most d, and (b) the maximum degree from a vertex to some other part is at most mu, there is guaranteed to be a transversal of the parts that forms an independent set of H. This is a common strengthening of two results of Loh and Sudakov (2007) and Molloy and Thron (2012), each of which in turn implies an earlier result of Reed and Sudakov (2002).

Title: Building and Interpreting Deep Similarity Models Abstract: Many learning algorithms such as kernel machines, nearest neighbors, clustering, or anomaly detection, are based on distances or similarities. Before similarities are used for training an actual machine learning model, we would like to verify that they are bound to meaningful patterns in the data. In this paper, we propose to make similarities interpretable by augmenting them with an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">explanation</i> . We develop BiLRP, a scalable and theoretically founded method to systematically decompose the output of an already trained deep similarity model on pairs of input features. Our method can be expressed as a composition of LRP explanations, which were shown in previous works to scale to highly nonlinear models. Through an extensive set of experiments, we demonstrate that BiLRP robustly explains complex similarity models, e.g., built on VGG-16 deep neural network features. Additionally, we apply our method to an open problem in digital humanities: detailed assessment of similarity between historical documents, such as astronomical tables. Here again, BiLRP provides insight and brings verifiability into a highly engineered and problem-specific similarity model.

Title: Entropy of tropical holonomic sequences Abstract: We introduce tropical holonomic sequences of a given order and calculate their entropy in case of the second order.

Title: Mechanistic Modeling of Longitudinal Shape Changes: Equations of Motion and Inverse Problems Abstract: This paper examines a longitudinal shape evolution model in which a three-dimensional volume progresses through a family of elastic equilibria in response to the time-derivative of an internal force, or yank, with an additional regularization to ensure diffeomorphic transformations. We consider two different models of yank and address the long time existence and uniqueness of solutions for the equations of motion in both models. In addition, we derive sufficient conditions for the existence of an optimal yank that best describes the change from an observed initial volume to an observed volume at a later time. The main motivation for this work is the understanding of processes such as growth and atrophy in anatomical structures, where the yank could be roughly interpreted as a metabolic event triggering morphological changes. We provide preliminary results on simple examples to illustrate, under this model, the retrievability of some attributes of such events.

Title: On Hop-Constrained Steiner Trees in Tree-Like Metrics Abstract: We consider the problem of computing a Steiner tree of minimum cost under a hop constraint which requires the depth of the tree to be at most $k$. Our main result is an exact algorithm for metrics induced by graphs with bounded treewidth that runs in time $n^{O(k)}$. For the special case of a path, we give a simple algorithm that solves the problem in polynomial time, even if $k$ is part of the input. The main result can be used to obtain, in quasi-polynomial time, a near-optimal solution that violates the $k$-hop constraint by at most one hop for more general metrics induced by graphs of bounded highway dimension and bounded doubling dimension. For non-metric graphs, we rule out an $o(\log n)$-approximation, assuming P$\neq$NP even when relaxing the hop constraint by any additive constant.

Title: IG-RL: Inductive Graph Reinforcement Learning for Massive-Scale Traffic Signal Control Abstract: Scaling adaptive traffic signal control involves dealing with combinatorial state and action spaces. Multi-agent reinforcement learning attempts to address this challenge by distributing control to specialized agents. However, specialization hinders generalization and transferability, and the computational graphs underlying neural-network architectures-dominating in the multi-agent setting-do not offer the flexibility to handle an arbitrary number of entities which changes both between road networks, and over time as vehicles traverse the network. We introduce Inductive Graph Reinforcement Learning (IG-RL) based on graph-convolutional networks which adapts to the structure of any road network, to learn detailed representations of traffic signal controllers and their surroundings. Our decentralized approach enables learning of a transferable-adaptive-trafficsignal-control policy. After being trained on an arbitrary set of road networks, our model can generalize to new road networks and traffic distributions, with no additional training and a constant number of parameters, enabling greater scalability compared to prior methods. Furthermore, our approach can exploit the granularity of available data by capturing the (dynamic) demand at both the lane level and the vehicle level. The proposed method is tested on both road networks and traffic settings never experienced during training. We compare IG-RL to multi-agent reinforcement learning and domain-specific baselines. In both synthetic road networks and in a larger experiment involving the control of the 3,971 traffic signals of Manhattan, we show that different instantiations of IG-RL outperform baselines.

Title: Level set and density estimation on manifolds Abstract: We tackle the problem of the estimation of the level sets Lf(λ) of the density f of a random vector X supported on a smooth manifold M⊂Rd, from an iid sample of X. To do that we introduce a kernel-based estimator fˆn,h, which is a slightly modified version of the one proposed in Rodríguez-Casal and Saavedra-Nieves (2014) and proves its a.s. uniform convergence to f. Then, we propose two estimators of Lf(λ), the first one is a plug-in: Lfˆn,h(λ), which is proven to be a.s. consistent in Hausdorff distance and distance in measure, if Lf(λ) does not meet the boundary of M. While the second one assumes that Lf(λ) is r-convex, and is estimated by means of the r-convex hull of Lfˆn,h(λ). The performance of our proposal is illustrated through some simulated examples. In a real data example we analyze the intensity and direction of strong and moderate winds.

Title: SARDO: An Automated Search-and-Rescue Drone-Based Solution for Victims Localization Abstract: Natural disasters affect millions of people every year. Finding missing persons in the shortest possible time is of crucial importance to reduce the death toll. This task is especially challenging when victims are sparsely distributed in large and/or difficult-to-reach areas and cellular networks are down. In this paper we present SARDO, a drone-based search and rescue solution that leverages the high penetration rate of mobile phones in the society to localize missing people. SARDO is an autonomous, all-in-one drone-based mobile network solution that does not require infrastructure support or mobile phones modifications. It builds on novel concepts such as pseudo-trilateration combined with machine-learning techniques to efficiently locate mobile phones in a given area. Our results, with a prototype implementation in a field-[1], show that SARDO rapidly determines the location of mobile phones ( <inline-formula><tex-math notation="LaTeX">$\sim \!3$</tex-math></inline-formula> min/UE) in a given area with an accuracy of few tens of meters and at a low battery consumption cost ( <inline-formula><tex-math notation="LaTeX">$\sim \!5\%$</tex-math></inline-formula> ). State-of-the-art localization solutions for disaster scenarios rely either on mobile infrastructure support or exploit onboard cameras for human/computer vision, IR, thermal-based localization. To the best of our knowledge, SARDO is the first drone-based cellular search-and-rescue solution able to accurately localize missing victims through mobile phones.

Title: On the Interpretation of the Balance Function Abstract: We construct a simple toy model and explicitly demonstrate that the balance function (BF) can become negative for some values of the rapidity separation and hence cannot have any probabilistic interpretation. In particular, the BF cannot be interpreted as the probability density for the balancing charges to occur separated by the given rapidity interval.

Title: Wasserstein-Based Graph Alignment Abstract: A novel method for comparing non-aligned graphs of various sizes is proposed, based on the Wasserstein distance between graph signal distributions induced by the respective graph Laplacian matrices. Specifically, a new formulation for the one-to-many graph alignment problem is casted, which aims at matching a node in the smaller graph with one or more nodes in the larger graph. By incorporating optimal transport into our graph comparison framework, a structurally-meaningful graph distance, and a signal transportation plan that models the structure of graph data are generated. The resulting alignment problem is solved with stochastic gradient descent, where a novel Dykstra operator is used to ensure that the solution is a one-to-many (soft) assignment matrix. The performance of our novel framework is demonstrated on graph alignment, graph classification and graph signal transportation. Our method is shown to lead to significant improvements with respect to the state-of-the-art algorithms on each ofthese tasks.

Title: Heterogeneous relational reasoning in knowledge graphs with reinforcement learning Abstract: Path-based relational reasoning over knowledge graphs has become increasingly popular due to a variety of downstream applications such as question answering in dialogue systems, fact prediction, and recommendation systems. In recent years, reinforcement learning (RL) based solutions for knowledge graphs have been demonstrated to be more interpretable and explainable than other deep learning models. However, the current solutions still struggle with performance issues due to incomplete state representations and large action spaces for the RL agent. We address these problems by developing HRRL (Heterogeneous Relational reasoning with Reinforcement Learning), a type-enhanced RL agent that utilizes the local heterogeneous neighborhood information for efficient path-based reasoning over knowledge graphs. HRRL improves the state representation using a graph neural network (GNN) for encoding the neighborhood information and utilizes entity type information for pruning the action space. Extensive experiments on real-world datasets show that HRRL outperforms state-of-the-art RL methods and discovers more novel paths during the training procedure, demonstrating the explorative power of our method.

Title: A comparison of parameter estimation in function-on-function regression Abstract: Recent technological developments have enabled us to collect complex and high-dimensional data in many scientific fields, such as population health, meteorology, econometrics, geology, and psychology. It is common to encounter such datasets collected repeatedly over a continuum. Functional data, whose sample elements are functions in the graphical forms of curves, images, and shapes, characterize these data types. Functional data analysis techniques reduce the complex structure of these data and focus on the dependences within and (possibly) between the curves. A common research question is to investigate the relationships in regression models that involve at least one functional variable. However, the performance of functional regression models depends on several factors, such as the smoothing technique, the number of basis functions, and the estimation method. This article provides a selective comparison for function-on-function regression models where both the response and predictor(s) are functions, to determine the optimal choice of basis function from a set of model evaluation criteria. We also propose a bootstrap method to construct a confidence interval for the response function. The numerical comparisons are implemented through Monte Carlo simulations and two real data examples.

Title: Targeting customers under response-dependent costs Abstract: •A formal decision analysis clarifies the economics of the customer targeting problem.•Optimal targeting considers the marketing effect and customer response probability.•Causal hurdle models jointly estimate treatment effect and response probability.•Results from e-coupon targeting confirm the superiority of the analytical policy.

Title: BayesFlow: Learning Complex Stochastic Models With Invertible Neural Networks Abstract: Estimating the parameters of mathematical models is a common problem in almost all branches of science. However, this problem can prove notably difficult when processes and model descriptions become increasingly complex and an explicit likelihood function is not available. With this work, we propose a novel method for globally amortized Bayesian inference based on invertible neural networks that w...

Title: Integral input-to-state stability of unbounded bilinear control systems Abstract: We study integral input-to-state stability of bilinear systems with unbounded control operators and derive natural sufficient conditions. The results are applied to a bilinearly controlled Fokker–Planck equation.

Title: Solvable Criterion for the Contextuality of any Prepare-and-Measure Scenario Abstract: Starting from arbitrary sets of quantum states and measurements, referred to as the prepare-and-measure scenario, an operationally noncontextual ontological model of the quantum statistics associated with the prepare-and-measure scenario is constructed. The operationally noncontextual ontological model coincides with standard Spekkens noncontextual ontological models for tomographically complete scenarios, while covering the non-tomographically complete case with a new notion of a reduced space, which we motivate following the guiding principles of noncontextuality. A mathematical criterion, called unit separability, is formulated as the relevant classicality criterion - the name is inspired by the usual notion of quantum state separability. Using this criterion, we derive a new upper bound on the cardinality of the ontic space. Then, we recast the unit separability criterion as a (possibly infinite) set of linear constraints, from which we obtain two separate hierarchies of algorithmic tests to witness the non-classicality or certify the classicality of a scenario. Finally, we reformulate our results in the framework of generalized probabilistic theories and discuss the implications for simplex-embeddability in such theories.

Title: Neural generators of sparse local linear models for achieving both accuracy and interpretability Abstract: For reliability, it is important for the predictions made by machine learning methods to be interpretable by humans. In general, deep neural networks (DNNs) can provide accurate predictions, although it is difficult to interpret why such predictions are obtained by the DNNs. On the other hand, interpretation of linear models is easy, although their predictive performance is low because real-world data are often intrinsically non-linear. To combine both the benefits of the high predictive performance of DNNs and the high interpretability of linear models into a single model, we propose neural generators of sparse local linear models (NGSLL). Sparse local linear models have high flexibility because they can approximate non-linear functions. NGSLL generates sparse linear weights for each sample using DNNs that take the original representations of each sample (e.g., word sequence) and their simplified representations (e.g., bag-of-words) as input. By extracting features from the original representations, the weights can contain rich information and achieve a high predictive performance. In addition, the prediction is interpretable because it is obtained through the inner product between the simplified representations and the sparse weights, where only a small number of weights are selected by our gate module in NGSLL. In experiments on image, text and tabular datasets, we demonstrate the effectiveness of NGSLL quantitatively and qualitatively by evaluating the prediction performance and visualizing generated weights.