Datasets:

yyu

/

arxiv-attrprompt

like 1

physics.flu-dyn physics.geo-ph

Universal fluctuations in the bulk of Rayleigh B 'enard turbulence. We present an investigation of the root mean square (rms) temperature sigma T and the rms velocity sigma w in the bulk of Rayleigh B 'enard turbulence, using new experimental data from the current study and experimental and numerical data from previous studies. We find that, once scaled by the convective temperature theta , the value of sigma T at the cell centre is a constant, i.e. sigma T,c theta approx 0.85 , over a wide range of the Rayleigh number ( 10 8 leq Ra leq 10 15 ) and the Prandtl number ( 0.7 leq Pr leq 23.34 ), and is independent of the surface topographies of the top and bottom plates of the convection cell. A constant close to unity suggests that theta is a proper measure of the temperature fluctuation in the core region. On the other hand, sigma w,c w , the vertical rms velocity at the cell centre scaled by the convective velocity w , shows a weak Ra dependence ( sim Ra 0.07 pm0.02 ) over 10 8 leq Ra leq 10 10 at Pr sim4.3 and is independent of plate topography. Similar to a previous finding by He Xia ( it Phys. Rev. Lett., vol. 122, 2019, 014503), we find that the rms temperature profile sigma T(z) theta in the region of the mixing zone with a mean horizontal shear exhibits a power law dependence on the distance z from the plate, but now the universal profile applies to both smooth and rough surface topographies and over a wider range of Ra . The vertical rms velocity profile sigma w(z) w obey a logarithmic dependence on z . The study thus demonstrates that the typical scales for the temperature and the velocity are the convective temperature theta and the the convective velocity w , respectively. Finally, we note that theta may be utilised to study the flow regime transitions in the ultra high Ra number turbulent convection.

math.AT math.GR

Group actions of A 5 on contractible 2 complexes. We prove that every action of A 5 on a finite 2 dimensional contractible complex has a fixed point.

cs.LO

Harnessing Higher Order (Meta )Logic to Represent and Reason with Complex Ethical Theories. The computer mechanization of an ambitious explicit ethical theory, Gewirth's Principle of Generic Consistency, is used to showcase an approach for representing and reasoning with ethical theories exhibiting complex logical features like alethic and deontic modalities, indexicals, higher order quantification, among others. Harnessing the high expressive power of Church's type theory as a meta logic to semantically embed a combination of quantified non classical logics, our work pushes existing boundaries in knowledge representation and reasoning. We demonstrate that intuitive encodings of complex ethical theories and their automation on the computer are no longer antipodes.

cs.DB cs.DC

Distributed Subtrajectory Join on Massive Datasets. Joining trajectory datasets is a significant operation in mobility data analytics and the cornerstone of various methods that aim to extract knowledge out of them. In the era of Big Data, the production of mobility data has become massive and, consequently, performing such an operation in a centralized way is not feasible. In this paper, we address the problem of Distributed Subtrajectory Join processing by utilizing the MapReduce programming model. Compared to traditional trajectory join queries, this problem is even more challenging since the goal is to retrieve all the "maximal" portions of trajectories that are "similar". We propose three solutions (i) a well designed basic solution, coined DTJb, (ii) a solution that uses a preprocessing step that repartitions the data, labeled DTJr, and (iii) a solution that, additionally, employs an indexing scheme, named DTJi. In our experimental study, we utilize a 56GB dataset of real trajectories from the maritime domain, which, to the best of our knowledge, is the largest real dataset used for experimentation in the literature of trajectory data management. The results show that DTJi performs up to 16x faster compared with DTJb, 10x faster than DTJr and 3x faster than the closest related state of the art algorithm.

math.AG cs.RO cs.SC

Hexapods with a small linear span. The understanding of mobile hexapods, i.e., parallel manipulators with six legs, is one of the driving questions in theoretical kinematics. We aim at contributing to this understanding by employing techniques from algebraic geometry. The set of configurations of a mobile hexapod with one degree of freedom has the structure of a projective curve, which hence has a degree and an embedding dimension. Our main result is a classification of configuration curves of hexapods that satisfy some restrictions on their embedding dimension.

physics.class-ph

Images of point charges in conducting ellipses and prolate spheroids. This is an investigation into the exact forms of images of point charges in 2D conducting ellipses and 3D prolate spheroids. For an ellipse with exterior point source, we compare two previous expressions for the analytic continuation inside the ellipse down to the focal line, the location of the image charge. For an interior source, we discover a sequence of image charges lying outside. For a point source close the surface of a spheroid, series solutions for the potential diverge in a region that encompasses the singularities of the continuation of the reflected potential inside the spheroid. We uncover an image system for a point charge on the axis of a prolate spheroid that extends from the focal segment somewhat further up the axis. For an off axis charge we find a new approximate image charge. For a point charge at the center of the spheroid, the image singularity is found to lie on an infinitely wide, flat sheet with a hole around the spheroid. For a point charge located anywhere inside the spheroid, we find two new approximate image charge solutions, a point charge in real space which applies when source lies near the surface, or a point charge in complex space which projects as a ring in the real space which applies when the source lies near the rotation axis. When the point charge lies exactly on a focal point, a multipole of order 1 2 is an ideal image approximation.

physics.plasm-ph physics.acc-ph physics.data-an

Beyond optimization supervised learning applications in relativistic laser plasma experiments. We explore the applications of machine learning techniques in relativistic laser plasma experiments beyond optimization purposes. We predict the beam charge of electrons produced in a laser wakefield accelerator given the laser wavefront change caused by a deformable mirror. Machine learning enables feature analysis beyond merely searching for an optimal beam charge, showing that specific aberrations in the laser wavefront are favored in generating higher beam charges. Supervised learning models allow characterizing the measured data quality as well as recognizing irreproducible data and potential outliers. We also include virtual measurement errors in the experimental data to examine the model robustness under these conditions. This work demonstrates how machine learning methods can benefit data analysis and physics interpretation in a highly nonlinear problem of relativistic laser plasma interaction.

math.FA

Polynomial Valuations on Vector Lattices. We prove that polynomial valuations on vector lattices correspond to orthosymmetric multilinear maps. As a consequence we obtain a concise proof of the equivalence of orthosymmetry and orthogonal additivity.

cs.SI cs.AI

Refining Similarity Matrices to Cluster Attributed Networks Accurately. As a result of the recent popularity of social networks and the increase in the number of research papers published across all fields, attributed networks consisting of relationships between objects, such as humans and the papers, that have attributes are becoming increasingly large. Therefore, various studies for clustering attributed networks into sub networks are being actively conducted. When clustering attributed networks using spectral clustering, the clustering accuracy is strongly affected by the quality of the similarity matrices, which are input into spectral clustering and represent the similarities between pairs of objects. In this paper, we aim to increase the accuracy by refining the matrices before applying spectral clustering to them. We verify the practicability of our proposed method by comparing the accuracy of spectral clustering with similarity matrices before and after refining them.

math.GR

Recognizing finite matrix groups over infinite fields. We present a uniform methodology for computing with finitely generated matrix groups over any infinite field. As one application, we completely solve the problem of deciding finiteness in this class of groups. We also present an algorithm that, given such a finite group as input, in practice successfully constructs an isomorphic copy over a finite field, and uses this copy to investigate the group's structure. Implementations of our algorithms are available in MAGMA.

cs.CL cs.LG cs.SI

Two path Deep Semi supervised Learning for Timely Fake News Detection. News in social media such as Twitter has been generated in high volume and speed. However, very few of them are labeled (as fake or true news) by professionals in near real time. In order to achieve timely detection of fake news in social media, a novel framework of two path deep semi supervised learning is proposed where one path is for supervised learning and the other is for unsupervised learning. The supervised learning path learns on the limited amount of labeled data while the unsupervised learning path is able to learn on a huge amount of unlabeled data. Furthermore, these two paths implemented with convolutional neural networks (CNN) are jointly optimized to complete semi supervised learning. In addition, we build a shared CNN to extract the low level features on both labeled data and unlabeled data to feed them into these two paths. To verify this framework, we implement a Word CNN based semi supervised learning model and test it on two datasets, namely, LIAR and PHEME. Experimental results demonstrate that the model built on the proposed framework can recognize fake news effectively with very few labeled data.

math.LO math.GR

Sets, groups, and fields definable in vector spaces with a bilinear form. We study definable sets, groups, and fields in the theory T infty of infinite dimensional vector spaces over an algebraically closed field equipped with a nondegenerate symmetric (or alternating) bilinear form. First, we define an ( mathbb N times mathbb Z , leq lex ) valued dimension on definable sets in T infty enjoying many properties of Morley rank in strongly minimal theories. Then, using this dimension notion as the main tool, we prove that all groups definable in T infty are (algebraic by abelian) by algebraic, which, in particular, answers a question of Granger. We conclude that every infinite field definable in T infty is definably isomorphic to the field of scalars of the vector space. We derive some other consequences of good behaviour of the dimension in T infty , e.g. every generic type in any definable set is a definable type every set is an extension base every definable group has a definable connected component. We also consider the theory T RCF infty of vector spaces over a real closed field equipped with a nondegenerate alternating bilinear form or a nondegenerate symmetric positive definite bilinear form. Using the same construction as in the case of T infty , we define a dimension on sets definable in T RCF infty , and using it we prove analogous results about definable groups and fields every group definable in T RCF infty is (semialgebraic by abelian) by semialgebraic (in particular, it is (Lie by abelian) by Lie), and every field definable in T RCF infty is definable in the field of scalars, hence it is either real closed or algebraically closed.

cs.CY

Governance of the Internet of Things (IoT). Today's increasing rate of technological change results from the rapid growth in computer processing speed, when combined with the cost decline of processing capacity, and is of historical import. The daily life of billions of individuals worldwide has been forever changed by technology in just the last few years. Costly data breaches continue at an alarming rate. The challenge facing humans as they attempt to govern the process of artificial intelligence, machine learning, and the impact of billions of sensory devices connected to the Internet is the subject of this Article. We proceed in nine sections. First, we define the Internet of Things (IoT), comment on the explosive growth in sensory devices connected to the Internet, provide examples of IoT devices, and speak to the promise of the IoT. Second, we discuss legal requirements for corporate governance as a foundation for considering the challenge of governing the IoT. Third, we look at potential IoT threats. Fourth, we discuss the Mirai botnet. Fifth, is a look at the IoT threat vector vulnerabilities during times of crisis. Sixth, we discuss the Manufactured Usage Description (MUD) methodology. Seventh, is a discussion of recent regulatory developments. Next, we look at a few recommendations. And finally, we conclude. We believe this Article contributes to our understanding of the widespread exposure to malware associated with IoT and adds to the nascent but emerging literature on governance of enterprise risk, a subject of vital societal importance.

math.OC cs.NA math.NA

Optimization and Learning With Nonlocal Calculus. Nonlocal models have recently had a major impact in nonlinear continuum mechanics and are used to describe physical systems processes which cannot be accurately described by classical, calculus based "local" approaches. In part, this is due to their multiscale nature that enables aggregation of micro level behavior to obtain a macro level description of singular irregular phenomena such as peridynamics, crack propagation, anomalous diffusion and transport phenomena. At the core of these models are nonlocal differential operators, including nonlocal analogs of the gradient Hessian. This paper initiates the use of such nonlocal operators in the context of optimization and learning. We define and analyze the convergence properties of nonlocal analogs of (stochastic) gradient descent and Newton's method on Euclidean spaces. Our results indicate that as the nonlocal interactions become less noticeable, the optima corresponding to nonlocal optimization converge to the "usual" optima. At the same time, we argue that nonlocal learning is possible in situations where standard calculus fails. As a stylized numerical example of this, we consider the problem of non differentiable parameter estimation on a non smooth translation manifold and show that our nonlocal gradient descent recovers the unknown translation parameter from a non differentiable objective function.

cs.SI cs.IT cs.LG math.IT q-bio.GN

ComHapDet A Spatial Community Detection Algorithm for Haplotype Assembly. Background Haplotypes, the ordered lists of single nucleotide variations that distinguish chromosomal sequences from their homologous pairs, may reveal an individual's susceptibility to hereditary and complex diseases and affect how our bodies respond to therapeutic drugs. Reconstructing haplotypes of an individual from short sequencing reads is an NP hard problem that becomes even more challenging in the case of polyploids. While increasing lengths of sequencing reads and insert sizes color black helps improve accuracy of reconstruction , it also exacerbates computational complexity of the haplotype assembly task. This has motivated the pursuit of algorithmic frameworks capable of accurate yet efficient assembly of haplotypes from high throughput sequencing data. Results We propose a novel graphical representation of sequencing reads and pose the haplotype assembly problem as an instance of community detection on a spatial random graph. To this end, we construct a graph where each read is a node with an unknown community label associating the read with the haplotype it samples. Haplotype reconstruction can then be thought of as a two step procedure first, one recovers the community labels on the nodes (i.e., the reads), and then uses the estimated labels to assemble the haplotypes. Based on this observation, we propose ComHapDet a novel assembly algorithm for diploid and ployploid haplotypes which allows both bialleleic and multi allelic variants. Conclusions Performance of the proposed algorithm is benchmarked on simulated as well as experimental data obtained by sequencing Chromosome 5 of tetraploid biallelic emph Solanum Tuberosum (Potato). The results demonstrate the efficacy of the proposed method and that it compares favorably with the existing techniques.

math.AP

Two regularized energy preserving finite difference methods for the logarithmic Klein Gordon equation. We present and analyze two regularized finite difference methods which preserve energy of the logarithmic Klein Gordon equation (LogKGE). In order to avoid singularity caused by the logarithmic nonlinearity of the LogKGE, we propose a regularized logarithmic Klein Gordon equation (RLogKGE) with a small regulation parameter 0 varepsilon ll1 to approximate the LogKGE with the convergence order O( varepsilon) . By adopting the energy method, the inverse inequality, and the cut off technique of the nonlinearity to bound the numerical solution, the error bound O(h 2 frac tau 2 varepsilon 2 ) of the two schemes with the mesh size h , the time step tau and the parameter varepsilon . Numerical results are reported to support our conclusions.

cs.CV

GREEN a Graph REsidual rE ranking Network for Grading Diabetic Retinopathy. The automatic grading of diabetic retinopathy (DR) facilitates medical diagnosis for both patients and physicians. Existing researches formulate DR grading as an image classification problem. As the stages categories of DR correlate with each other, the relationship between different classes cannot be explicitly described via a one hot label because it is empirically estimated by different physicians with different outcomes. This class correlation limits existing networks to achieve effective classification. In this paper, we propose a Graph REsidual rE ranking Network (GREEN) to introduce a class dependency prior into the original image classification network. The class dependency prior is represented by a graph convolutional network with an adjacency matrix. This prior augments image classification pipeline by re ranking classification results in a residual aggregation manner. Experiments on the standard benchmarks have shown that GREEN performs favorably against state of the art approaches.

cs.DL cs.IR

Practical method to reclassify Web of Science articles into unique subject categories and broad disciplines. Classification of bibliographic items into subjects and disciplines in large databases is essential for many quantitative science studies. The Web of Science classification of journals into 250 subject categories, which has served as a basis for many studies, is known to have some fundamental problems and several practical limitations that may affect the results from such studies. Here we present an easily reproducible method to perform reclassification of the Web of Science into existing subject categories and into 14 broad areas. Our reclassification is at a level of articles, so it preserves disciplinary differences that may exist among individual articles published in the same journal. Reclassification also eliminates ambiguous (multiple) categories that are found for 50 of items, and assigns a discipline field category to all articles that come from broad coverage journals such as Nature and Science. The correctness of the assigned subject categories is evaluated manually and is found to be 95 .

cs.CR cs.DC

The Security Reference Architecture for Blockchains Towards a Standardized Model for Studying Vulnerabilities, Threats, and Defenses. Blockchains are distributed systems, in which security is a critical factor for their success. However, despite their increasing popularity and adoption, there is a lack of standardized models that study blockchain related security threats. To fill this gap, the main focus of our work is to systematize and extend the knowledge about the security and privacy aspects of blockchains and contribute to the standardization of this domain. We propose the security reference architecture (SRA) for blockchains, which adopts a stacked model (similar to the ISO OSI) describing the nature and hierarchy of various security and privacy aspects. The SRA contains four layers (1) the network layer, (2) the consensus layer, (3) the replicated state machine layer, and (4) the application layer. At each of these layers, we identify known security threats, their origin, and countermeasures, while we also analyze several cross layer dependencies. Next, to enable better reasoning about security aspects of blockchains by the practitioners, we propose a blockchain specific version of the threat risk assessment standard ISO IEC 15408 by embedding the stacked model into this standard. Finally, we provide designers of blockchain platforms and applications with a design methodology following the model of SRA and its hierarchy.

math.OA

Nuclearity of semigroup C algebras. We study the semigroup C algebra of a positive cone P of a weakly quasi lattice ordered group. That is, P is a subsemigroup of a discrete group G with P cap P 1 e and such that any two elements of P with a common upper bound in P also have a least upper bound. We find sufficient conditions for the semigroup C algebra of P to be nuclear. These conditions involve the idea of a generalised length function, called a "controlled map", into an amenable group. Here we give a new definition of a controlled map and discuss examples from different sources. We apply our main result to establish nuclearity for semigroup C algebras of a class of one relator semigroups, motivated by a recent work of Li, Omland and Spielberg. This includes all the Baumslag Solitar semigroups. We also analyse semidirect products of weakly quasi lattice ordered groups and use our theorem in examples to prove nuclearity of the semigroup C algebra. Moreover, we prove that the graph product of weak quasi lattices is again a weak quasi lattice, and show that the corresponding semigroup C algebra is nuclear when the underlying groups are amenable.

physics.app-ph

Direct Imaging of Resonant Phonon Magnon Coupling. Direct detection of phonons is important for the investigation of information interconversion between the resonantly coupled magnons and phonons. Here we report resonant coupling of magnons and phonons, which can be directly visualized by using micro focused Brillouin light scattering in Ni LiNbO3 hybrid heterostructures. The patterns of surface acoustic wave phonons, originating from the interference between the original wave psi 0(A 0,k) and reflected wave psi 1(A 1, k), can be modulated by magnetic field due to the magnon phonon coupling. By analyzing the information of phonons obtained from Brillouin spectroscopy, the properties of the magnon system (Ni film), e.g., ferromagnetic resonance field and resonance linewidth can be determined. The results provide spatially resolved information about phonon manipulation and detection in a coupled magnon phonon system.

cs.DS cs.NA cs.SI math.NA

GRASS Graph Spectral Sparsification Leveraging Scalable Spectral Perturbation Analysis. Spectral graph sparsification aims to find ultra sparse subgraphs whose Laplacian matrix can well approximate the original Laplacian eigenvalues and eigenvectors. In recent years, spectral sparsification techniques have been extensively studied for accelerating various numerical and graph related applications. Prior nearly linear time spectral sparsification methods first extract low stretch spanning tree from the original graph to form the backbone of the sparsifier, and then recover small portions of spectrally critical off tree edges to the spanning tree to significantly improve the approximation quality. However, it is not clear how many off tree edges should be recovered for achieving a desired spectral similarity level within the sparsifier. Motivated by recent graph signal processing techniques, this paper proposes a similarity aware spectral graph sparsification framework that leverages efficient spectral off tree edge embedding and filtering schemes to construct spectral sparsifiers with guaranteed spectral similarity (relative condition number) level. An iterative graph densification scheme is also introduced to facilitate efficient and effective filtering of off tree edges for highly ill conditioned problems. The proposed method has been validated using various kinds of graphs obtained from public domain sparse matrix collections relevant to VLSI CAD, finite element analysis, as well as social and data networks frequently studied in many machine learning and data mining applications. For instance, a sparse SDD matrix with 40 million unknowns and 180 million nonzeros can be solved (1E 3 accuracy level) within two minutes using a single CPU core and about 6GB memory.

math.CA math.CV math.FA math.SP

New solution of a problem of Kolmogorov on width asymptotics in holomorphic function spaces. Given a domain D in mathbb C n and K a compact subset of D , the set mathcal A K D of all restrictions of functions holomorphic on D the modulus of which is bounded by 1 is a compact subset of the Banach space C(K) of continuous functions on K . The sequence (d m( mathcal A K D)) m in mathbb N of Kolmogorov m widths of mathcal A K D provides a measure of the degree of compactness of the set mathcal A K D in C(K) and the study of its asymptotics has a long history, essentially going back to Kolmogorov's work on epsilon entropy of compact sets in the 1950s. In the 1980s Zakharyuta showed that for suitable D and K the asymptotics begin equation lim m to infty frac log d m( mathcal A K D) m 1 n 2 pi left ( frac n! C(K,D) right ) 1 n ,, end equation where C(K,D) is the Bedford Taylor relative capacity of K in D is implied by a conjecture, now known as Zakharyuta's Conjecture, concerning the approximability of the regularised relative extremal function of K and D by certain pluricomplex Green functions. Zakharyuta's Conjecture was proved by Nivoche in 2004 thus settling the asymptotics above at the same time. We shall give a new proof of the asymptotics above for D strictly hyperconvex and K non pluripolar which does not rely on Zakharyuta's Conjecture. Instead we proceed more directly by a two pronged approach establishing sharp upper and lower bounds for the Kolmogorov widths. The lower bounds follow from concentration results of independent interest for the eigenvalues of a certain family of Toeplitz operators, while the upper bounds follow from an application of the Bergman Weil formula together with an exhaustion procedure by special holomorphic polyhedra.

cs.CV cs.LG

Learning Furniture Compatibility with Graph Neural Networks. We propose a graph neural network (GNN) approach to the problem of predicting the stylistic compatibility of a set of furniture items from images. While most existing results are based on siamese networks which evaluate pairwise compatibility between items, the proposed GNN architecture exploits relational information among groups of items. We present two GNN models, both of which comprise a deep CNN that extracts a feature representation for each image, a gated recurrent unit (GRU) network that models interactions between the furniture items in a set, and an aggregation function that calculates the compatibility score. In the first model, a generalized contrastive loss function that promotes the generation of clustered embeddings for items belonging to the same furniture set is introduced. Also, in the first model, the edge function between nodes in the GRU and the aggregation function are fixed in order to limit model complexity and allow training on smaller datasets in the second model, the edge function and aggregation function are learned directly from the data. We demonstrate state of the art accuracy for compatibility prediction and "fill in the blank" tasks on the Bonn and Singapore furniture datasets. We further introduce a new dataset, called the Target Furniture Collections dataset, which contains over 6000 furniture items that have been hand curated by stylists to make up 1632 compatible sets. We also demonstrate superior prediction accuracy on this dataset.

cs.CR cs.SI

Privacy protocols. Security protocols enable secure communication over insecure channels. Privacy protocols enable private interactions over secure channels. Security protocols set up secure channels using cryptographic primitives. Privacy protocols set up private channels using secure channels. But just like some security protocols can be broken without breaking the underlying cryptography, some privacy protocols can be broken without breaking the underlying security. Such privacy attacks have been used to leverage e commerce against targeted advertising from the outset but their depth and scope became apparent only with the overwhelming advent of influence campaigns in politics. The blurred boundaries between privacy protocols and privacy attacks present a new challenge for protocol analysis. Covert channels turn out to be concealed not only below overt channels, but also above subversions, and the level below attacks are supplemented by sublimations and the level above attacks.

math.AP

Finite time stability for the Riemann problem with extremal shocks for a large class of hyperbolic systems. In this paper on hyperbolic systems of conservation laws in one space dimension, we give a complete picture of stability for all solutions to the Riemann problem which contain only extremal shocks. We study stability of the Riemann problem amongst a large class of solutions. We show stability among the family of solutions with shocks from any family. We assume solutions verify at least one entropy condition. We have no small data assumptions. The solutions we consider are bounded and satisfy a strong trace condition weaker than BV text loc . We make only mild assumptions on the system. In particular, our work applies to gas dynamics, including the isentropic Euler system and the full Euler system for a polytropic gas. We use the theory of a contraction (see Kang and Vasseur Arch. Ration. Mech. Anal., 222(1) 343 391, 2016 ), and introduce new ideas in this direction to allow for two shocks from different shock families to be controlled simultaneously. This paper shows L 2 stability for the Riemann problem for all time. Our results compare to Chen, Frid, and Li Comm. Math. Phys., 228(2) 201 217, 2002 and Chen and Li J. Differential Equations, 202(2) 332 353, 2004 , which give uniqueness and long time stability for perturbations of the Riemann problem amongst a large class of solutions without smallness assumptions and which are locally BV . Although, these results lack global L 2 stability.

cs.DC cs.DB cs.SE

A Link Generator for Increasing the Utility of OpenAPI to GraphQL Translations. Standardized interfaces are the connecting link of today's distributed systems, facilitating access to data services in the cloud. REST APIs have been prevalent over the last years, despite several issues like over and underfetching of resources. GraphQL enjoys rapid adoption, resolving these problems by using statically typed queries. However, the redevelopment of services to the new paradigm is costly. Therefore, several approaches for the successive migration from REST to GraphQL have been proposed, many leveraging OpenAPI service descriptions. In this article, we present the findings of our empirical evaluation on the APIs.guru directory and identify several schema translation challenges. These include less expressive schema types in GraphQL, as well as missing meta information about related resources in OpenAPI. To this end, we developed the open source Link Generator, that analyzes OpenAPI documents and automatically adds links to increase translation utility. This fundamentally benefits around 34 of APIs in the APIs.guru directory. Our findings and tool support contribute to the ongoing discussion about the migration of REST APIs to GraphQL, and provide developers with valuable insights into common pitfalls, to reduce friction during API transformation.

physics.app-ph

Grazing incidence X ray fluorescence for a dimensional and elemental characterization of well ordered nanostructures. The increasing importance of well controlled ordered nanostructures on surfaces represents a challenge for existing metrology techniques. To develop such nanostructures and monitor complex processing constraints fabrication, both a dimensional reconstruction of nanostructures and a characterization (ideally a quantitative characterization) of their composition is required. In this work, we present a soft X ray fluorescence based methodology that allows both of these requirements to be addressed at the same time. By applying the grazing incidence X ray fluorescence technique and thus utilizing the X ray standing wave field effect, nanostructures can be investigated with a high sensitivity with respect to their dimensional and compositional characteristics. By varying the incident angles of the exciting radiation, element sensitive fluorescence radiation is emitted from different regions inside the nanoobjects. By applying an adequate modeling scheme, these datasets can be used to determine the nanostructure characteristics. We demonstrate these capabilities by performing an element sensitive reconstruction of a lamellar grating made of Si 3 N 4 , where GIXRF data for the O K alpha and N K alpha fluorescence emission allows a thin oxide layer to be reconstructed on the surface of the grating structure. In addition, we employ the technique also to three dimensional nanostructures and derive both dimensional and compositional parameters in a quantitative manner.

physics.comp-ph

A Method to Apply Piola Kirchhoff Stress in Molecular Statics Simulation. A force based optimization method is proposed to apply the first and second kind of Piola Kirchhoff stresses in molecular statics simulation. This method is important for finite deformation problems in which the atomistic behavior can be more accurately described using Piola Kirchhoff stresses. The performance of the method is tested and validated using Silicon as a model material.

cs.DC cs.DS cs.NI

Scheduling Coflows with Dependency Graph. Applications in data parallel computing typically consist of multiple stages. In each stage, a set of intermediate parallel data flows (Coflow) is produced and transferred between servers to enable starting of next stage. While there has been much research on scheduling isolated coflows, the dependency between coflows in multi stage jobs has been largely ignored. In this paper, we consider scheduling coflows of multi stage jobs represented by general DAGs (Directed Acyclic Graphs) in a shared data center network, so as to minimize the total weighted completion time of jobs. This problem is significantly more challenging than the traditional coflow scheduling, as scheduling even a single multi stage job to minimize its completion time is shown to be NP hard. In this paper, we propose a polynomial time algorithm with approximation ratio of O( mu log(m) log( log(m))) , where mu is the maximum number of coflows in a job and m is the number of servers. For the special case that the jobs' underlying dependency graphs are rooted trees, we modify the algorithm and improve its approximation ratio. To verify the performance of our algorithms, we present simulation results using real traffic traces that show up to 53 improvement over the prior approach. We conclude the paper by providing a result concerning an optimality gap for scheduling coflows with general DAGs.

cs.RO cs.CV

Depth Ranging Performance Evaluation and Improvement for RGB D Cameras on Field Based High Throughput Phenotyping Robots. RGB D cameras have been successfully used for indoor High ThroughpuT Phenotyping (HTTP). However, their capability and feasibility for in field HTTP still need to be evaluated, due to the noise and disturbances generated by unstable illumination, specular reflection, and diffuse reflection, etc. To solve these problems, we evaluated the depth ranging performances of two consumer level RGB D cameras (RealSense D435i and Kinect V2) under in field HTTP scenarios, and proposed a strategy to compensate the depth measurement error. For performance evaluation, we focused on determining their optimal ranging areas for different crop organs. Based on the evaluation results, we proposed a brightness and distance based Support Vector Regression Strategy, to compensate the ranging error. Furthermore, we analyzed the depth filling rate of two RGB D cameras under different lighting intensities. Experimental results showed that 1) For RealSense D435i, its effective ranging area is 0.160, 1.400 m, and in field filling rate is approximately 90 . 2) For Kinect V2, it has a high ranging accuracy in the 0.497, 1.200 m, but its in field filling rate is less than 24.9 . 3) Our error compensation model can effectively reduce the influences of lighting intensity and target distance. The maximum MSE and minimum R2 of this model are 0.029 and 0.867, respectively. To sum up, RealSense D435i has better ranging performances than Kinect V2 on in field HTTP.

physics.soc-ph cs.MA

Modeling indoor level non pharmaceutical interventions during the COVID 19 pandemic a pedestrian dynamics based microscopic simulation approach. Mathematical modeling of epidemic spreading has been widely adopted to estimate the threats of epidemic diseases (i.e., the COVID 19 pandemic) as well as to evaluate epidemic control interventions. The indoor place is considered to be a significant epidemic spreading risk origin, but existing widely used epidemic spreading models are usually limited for indoor places since the dynamic physical distance changes between people are ignored, and the empirical features of the essential and non essential travel are not differentiated. In this paper, we introduce a pedestrian based epidemic spreading model that is capable of modeling indoor transmission risks of diseases during people's social activities. Taking advantage of the before and after mobility data from the University of Maryland COVID 19 Impact Analysis Platform, it's found that people tend to spend more time in grocery stores once their travel frequencies are restricted to a low level. In other words, an increase in dwell time could balance the decrease in travel frequencies and satisfy people's demand. Based on the pedestrian based model and the empirical evidence, combined non pharmaceutical interventions from different operational levels are evaluated. Numerical simulations show that restrictions on people's travel frequency and open hours of indoor places may not be universally effective in reducing average infection risks for each pedestrian who visit the place. Entry limitations can be a widely effective alternative, whereas the decision maker needs to balance the decrease in risky contacts and the increase in queue length outside the place that may impede people from fulfilling their travel needs.

cs.CL cs.LG

Multi Granular Text Encoding for Self Explaining Categorization. Self explaining text categorization requires a classifier to make a prediction along with supporting evidence. A popular type of evidence is sub sequences extracted from the input text which are sufficient for the classifier to make the prediction. In this work, we define multi granular ngrams as basic units for explanation, and organize all ngrams into a hierarchical structure, so that shorter ngrams can be reused while computing longer ngrams. We leverage a tree structured LSTM to learn a context independent representation for each unit via parameter sharing. Experiments on medical disease classification show that our model is more accurate, efficient and compact than BiLSTM and CNN baselines. More importantly, our model can extract intuitive multi granular evidence to support its predictions.

math.AP

Stability of contact lines in fluids 2D Navier Stokes flow. In this paper we study the dynamics of an incompressible viscous fluid evolving in an open top container in two dimensions. The fluid mechanics are dictated by the Navier Stokes equations. The upper boundary of the fluid is free and evolves within the container. The fluid is acted upon by a uniform gravitational field, and capillary forces are accounted for along the free boundary. The triple phase interfaces where the fluid, air above the vessel, and solid vessel wall come in contact are called contact points, and the angles formed at the contact point are called contact angles. The model that we consider integrates boundary conditions that allow for full motion of the contact points and angles. Equilibrium configurations consist of quiescent fluid within a domain whose upper boundary is given as the graph of a function minimizing a gravity capillary energy functional, subject to a fixed mass constraint. The equilibrium contact angles can take on any values between 0 and pi depending on the choice of capillary parameters. The main thrust of the paper is the development of a scheme of a priori estimates that show that solutions emanating from data sufficiently close to the equilibrium exist globally in time and decay to equilibrium at an exponential rate.

math.SP math.PR

The Microlocal Irregularity of Gaussian Noise. The study of random Fourier series, linear combinations of trigonometric functions whose coefficients are independent (in our case Gaussian) random variables with polynomially bounded means and standard deviations, dates back to Norbert Wiener in one of the original constructions of Brownian motion. A geometric generalization relevant e.g. to Euclidean quantum field theory with an infrared cutoff is the study of random Gaussian linear combinations of the eigenfunctions of the Laplace Beltrami operator on an arbitrary compact Riemannian manifold (M,g) , Gaussian noise Phi . I will prove that, when our random coefficients are independent Gaussians whose standard deviations obey polynomial asymptotics and whose means obey a corresponding polynomial upper bound, the resultant random mathscr H s wavefront set operatorname WF s( Phi) (defined as a subset of the cosphere bundle mathbb S M ) is either almost surely empty or almost surely the entirety of mathbb S M , depending on s in mathbb R , and we will compute the threshold s and the behavior of the wavefront set at it. Consequently, the random C infty wavefront set operatorname WF ( Phi) is almost surely the entirety of the cosphere bundle. The method of proof is as follows using Sazonov's theorem and its converse, it suffices to understand which compositions of microlocal cutoffs and inclusions of L 2 based fractional order Sobolev spaces are Hilbert Schmidt (HS), and the answer follows from general facts about the HS norms of the elements of the pseudodifferential calculus of Kohn and Nirenberg.

math.CA math.AP

On the refined Strichartz estimates. We present a slightly simpler proof of the multilinear refined Strichartz estimate, and prove a slightly more general linear refined Strichartz estimate. Our arguments seek to clarify the connection between these estimates, refined decoupling and tube incidences.

math.NA cs.NA

Structure exploiting methods for fast uncertainty quantification in multiphase flow through heterogeneous media. We present a computational framework for dimension reduction and surrogate modeling to accelerate uncertainty quantification in computationally intensive models with high dimensional inputs and function valued outputs. Our driving application is multiphase flow in saturated unsaturated porous media in the context of radioactive waste storage. For fast input dimension reduction, we utilize an approximate global sensitivity measure, for function value outputs, motivated by ideas from the active subspace methods. The proposed approach does not require expensive gradient computations. We generate an efficient surrogate model by combining a truncated Karhunen Lo ' e ve (KL) expansion of the output with polynomial chaos expansions, for the output KL modes, constructed in the reduced parameter space. We demonstrate the effectiveness of the proposed surrogate modeling approach with a comprehensive set of numerical experiments, where we consider a number of function valued (temporally or spatially distributed) QoIs.

cs.SI cs.LG

Ising Based Louvain Method Clustering Large Graphs with Specialized Hardware. Recent advances in specialized hardware for solving optimization problems such quantum computers, quantum annealers, and CMOS annealers give rise to new ways for solving real word complex problems. However, given current and near term hardware limitations, the number of variables required to express a large real world problem easily exceeds the hardware capabilities, thus hybrid methods are usually developed in order to utilize the hardware. In this work, we advocate for the development of hybrid methods that are built on top of the frameworks of existing state of art heuristics, thereby improving these methods. We demonstrate this by building on the so called Louvain method, which is one of the most popular algorithms for the Community detection problem and develop and Ising based Louvain method. The proposed method outperforms two state of the art community detection algorithms in clustering several small to large scale graphs. The results show promise in adapting the same optimization approach to other unsupervised learning heuristics to improve their performance.

math.CO math.AC

On the determinant of multiplication map of a monomial complete intersection ring. In this article, we consider the monomial complete intersection algebra mathbb K x,y langle x d,y q rangle in two variables. For elements l 1, ldots,l d q 2k of degree 1 , we give a formula of the deteminant of linear map from the homogeneous component of degree k to the homogenous component of degree d q k defined by the multiplication of l 1 cdots l d q 2k .

cs.MA cs.LO

Natural Strategic Abilities in Voting Protocols. Security properties are often focused on the technological side of the system. One implicitly assumes that the users will behave in the right way to preserve the property at hand. In real life, this cannot be taken for granted. In particular, security mechanisms that are difficult and costly to use are often ignored by the users, and do not really defend the system against possible attacks. Here, we propose a graded notion of security based on the complexity of the user's strategic behavior. More precisely, we suggest that the level to which a security property varphi is satisfied can be defined in terms of (a) the complexity of the strategy that the voter needs to execute to make varphi true, and (b) the resources that the user must employ on the way. The simpler and cheaper to obtain varphi , the higher the degree of security. We demonstrate how the idea works in a case study based on an electronic voting scenario. To this end, we model the vVote implementation of the Pret voting protocol for coercion resistant and voter verifiable elections. Then, we identify "natural" strategies for the voter to obtain receipt freeness, and measure the voter's effort that they require. We also look at how hard it is for the coercer to compromise the election through a randomization attack.

math.AP math.FA math.PR

Time regularity for local weak solutions of the heat equation on local Dirichlet spaces. We study the time regularity of local weak solutions of the heat equation in the context of local regular symmetric Dirichlet spaces. Under two basic and rather minimal assumptions, namely, the existence of certain cut off functions and a very weak L 2 Gaussian type upper bound for the heat semigroup, we prove that the time derivatives of a local weak solution of the heat equation are themselves local weak solutions. This applies, for instance, to local weak solutions of parabolic equations with uniformly elliptic symmetric divergence form second order operators with measurable coefficients. We describe some applications to the structure of ancient local weak solutions of such equations which generalize recent results of 8 and 33 .

cs.LG cs.AI

Distributed representation of patients and its use for medical cost prediction. Efficient representation of patients is very important in the healthcare domain and can help with many tasks such as medical risk prediction. Many existing methods, such as diagnostic Cost Groups (DCG), rely on expert knowledge to build patient representation from medical data, which is resource consuming and non scalable. Unsupervised machine learning algorithms are a good choice for automating the representation learning process. However, there is very little research focusing on onpatient level representation learning directly from medical claims. In this paper, weproposed a novel patient vector learning architecture that learns high quality,fixed length patient representation from claims data. We conducted several experiments to test the quality of our learned representation, and the empirical results show that our learned patient vectors are superior to vectors learned through other methods including a popular commercial model. Lastly, we provide potential clinical interpretation for using our representation on predictive tasks, as interpretability is vital in the healthcare domain

math.AP

The inviscid limit for the Navier Stokes equations with data analytic only near the boundary. We address the inviscid limit for the Navier Stokes equations in a half space, with initial datum that is analytic only close to the boundary of the domain, and has finite Sobolev regularity in the complement. We prove that for such data the solution of the Navier Stokes equations converges in the vanishing viscosity limit to the solution of the Euler equation, on a constant time interval.

math.OC

Convergence of proximal solutions for evolution inclusions with time dependent maximal monotone operators. This article studies the solutions of time dependent differential inclusions which is motivated by their utility in the modeling of certain physical systems. The differential inclusion is described by a time dependent set valued mapping having the property that, for a given time instant, the set valued mapping describes a maximal monotone operator. Under certain mild assumptions on the regularity with respect to the time argument, we construct a sequence of functions parameterized by the sampling time that corresponds to the discretization of the continuous time system. Using appropriate tools from functional and variational analysis, this sequence is then shown to converge to the unique solution of the original differential inclusion. The result is applied to develop conditions for well posedness of differential equations interconnected with nonsmooth time dependent complementarity relations, using passivity of underlying dynamics (equivalently expressed in terms of linear matrix inequalities).

math.FA math.AP

Sharp subcritical and critical L p Hardy inequalities on the sphere. We prove sharp inequalities of Hardy type for functions in the Sobolev space W 1,p on the unit sphere mathbb S n 1 in mathbb R n . We achieve this in both the subcritical and critical cases. The method we use to show optimality takes into account all the constants involved in our inequalities.

cs.CV

Towards More Realistic Human Robot Conversation A Seq2Seq based Body Gesture Interaction System. This paper presents a novel system that enables intelligent robots to exhibit realistic body gestures while communicating with humans. The proposed system consists of a listening model and a speaking model used in corresponding conversational phases. Both models are adapted from the sequence to sequence (seq2seq) architecture to synthesize body gestures represented by the movements of twelve upper body keypoints. All the extracted 2D keypoints are firstly 3D transformed, then rotated and normalized to discard irrelevant information. Substantial videos of human conversations from Youtube are collected and preprocessed to train the listening and speaking models separately, after which the two models are evaluated using metrics of mean squared error (MSE) and cosine similarity on the test dataset. The tuned system is implemented to drive a virtual avatar as well as Pepper, a physical humanoid robot, to demonstrate the improvement on conversational interaction abilities of our method in practice.

physics.app-ph

Elastic Weyl points and surface arc states in 3D structures. The study of Weyl points in electronic systems has inspired many recent researches in classical systems such as photonic and acoustic lattices. Here we show how the Weyl physics can also inspire the design of novel elastic structures. We construct a single phase 3D structure, an analogue of the AA stacked honeycomb lattice, and predict the existence of Weyl points with opposite topological charges ( pm 1 ), elastic Fermi arcs, and the associated gapless topologically protected surface states. We employ full scale numerical simulations on the elastic 3D structure, and present a clear visualization of topological surface states that are directional and robust. Such designed lattices can pave the way for novel vibration control and energy harvesting on structures that are ubiquitous in many engineering applications.

cs.CR cs.NI

When Distributed Ledger Technology meets Internet of Things Benefits and Challenges. There is a growing interest from both the academia and industry to employ distributed ledger technology in the Internet of Things domain for addressing security related and performance challenges. Distributed ledger technology enables non trusted entities to communicate and reach consensus in a fully distributed manner through a cryptographically secure and immutable ledger. However, significant challenges arise mainly related to transaction processing speed and user privacy. This work explores the interplay between Internet of Things and distributed ledger technology, analysing the fundamental characteristics of this technology and discussing the related benefits and challenges.

math.CO math.AG math.GR

Tur 'an and Ramsey problems for alternating multilinear maps. Guided by the connections between hypergraphs and exterior algebras, we study Tur 'an and Ramsey type problems for alternating multilinear maps. This study lies at the intersection of combinatorics, group theory, and algebraic geometry, and has origins in the works of Lov 'asz (Proc. Sixth British Combinatorial Conf., 1977), Buhler, Gupta, and Harris (J. Algebra, 1987), and Feldman and Propp (Adv. Math., 1992). Our main result is a Ramsey theorem for alternating bilinear maps. Given s, t in mathbb N , s, t geq 2 , and an alternating bilinear map f V times V to U with dim(V) s cdot t 4 , we show that there exists either a dimension s subspace W leq V such that dim(f(W, W)) 0 , or a dimension t subspace W leq V such that dim(f(W, W)) binom t 2 . This result has natural group theoretic (for finite p groups) and geometric (for Grassmannians) implications, and leads to new Ramsey type questions for varieties of groups and Grassmannians.

cs.CV cs.LG

Big Transfer (BiT) General Visual Representation Learning. Transfer of pre trained representations improves sample efficiency and simplifies hyperparameter tuning when training deep neural networks for vision. We revisit the paradigm of pre training on large supervised datasets and fine tuning the model on a target task. We scale up pre training, and propose a simple recipe that we call Big Transfer (BiT). By combining a few carefully selected components, and transferring using a simple heuristic, we achieve strong performance on over 20 datasets. BiT performs well across a surprisingly wide range of data regimes from 1 example per class to 1M total examples. BiT achieves 87.5 top 1 accuracy on ILSVRC 2012, 99.4 on CIFAR 10, and 76.3 on the 19 task Visual Task Adaptation Benchmark (VTAB). On small datasets, BiT attains 76.8 on ILSVRC 2012 with 10 examples per class, and 97.0 on CIFAR 10 with 10 examples per class. We conduct detailed analysis of the main components that lead to high transfer performance.

cs.CV cs.LG

A Generalized Framework for Agglomerative Clustering of Signed Graphs applied to Instance Segmentation. We propose a novel theoretical framework that generalizes algorithms for hierarchical agglomerative clustering to weighted graphs with both attractive and repulsive interactions between the nodes. This framework defines GASP, a Generalized Algorithm for Signed graph Partitioning, and allows us to explore many combinations of different linkage criteria and cannot link constraints. We prove the equivalence of existing clustering methods to some of those combinations, and introduce new algorithms for combinations which have not been studied. An extensive comparison is performed to evaluate properties of the clustering algorithms in the context of instance segmentation in images, including robustness to noise and efficiency. We show how one of the new algorithms proposed in our framework outperforms all previously known agglomerative methods for signed graphs, both on the competitive CREMI 2016 EM segmentation benchmark and on the CityScapes dataset.

cs.LG cs.AI cs.NE

Instance Weighted Incremental Evolution Strategies for Reinforcement Learning in Dynamic Environments. Evolution strategies (ES), as a family of black box optimization algorithms, recently emerge as a scalable alternative to reinforcement learning (RL) approaches such as Q learning or policy gradient, and are much faster when many central processing units (CPUs) are available due to better parallelization. In this paper, we propose a systematic incremental learning method for ES in dynamic environments. The goal is to adjust previously learned policy to a new one incrementally whenever the environment changes. We incorporate an instance weighting mechanism with ES to facilitate its learning adaptation, while retaining scalability of ES. During parameter updating, higher weights are assigned to instances that contain more new knowledge, thus encouraging the search distribution to move towards new promising areas of parameter space. We propose two easy to implement metrics to calculate the weights instance novelty and instance quality. Instance novelty measures an instance's difference from the previous optimum in the original environment, while instance quality corresponds to how well an instance performs in the new environment. The resulting algorithm, Instance Weighted Incremental Evolution Strategies (IW IES), is verified to achieve significantly improved performance on challenging RL tasks ranging from robot navigation to locomotion. This paper thus introduces a family of scalable ES algorithms for RL domains that enables rapid learning adaptation to dynamic environments.

cs.NI

Robust Cellular Communications for Unmanned Aerial Vehicles in Maritime Search and Rescue. Unmanned aerial vehicles (UAV) are a promising technology for fast, large scale survey operations such as maritime search and rescue (SAR). However, providing reliable communications over long distances remains a challenge. Cellular technology such as Long Term Evolution (LTE) is designed to perform well in the presence of fast moving clients and highly dynamic channel conditions. Yet, its generic medium access mechanisms make it hard to support low latency uplink traffic, which is needed for real time command and control and similar applications. In this paper, we introduce an unmanned aerial system for maritime SAR, supported via an LTE data link. To mitigate drawbacks in uplink channel access, we develop a resource guaranteed scheme based on persistent scheduling, using an open source LTE stack. The approach is evaluated with a laboratory setup using software defined radio modules. Measurements are taken on the application layer, using real world telemetry data generated by an autopilot system in the presence of high bandwidth background traffic. Evaluations demonstrate that our system fulfills UAV requirements in terms of latency and reliability. Latency is further reduced to 6 ms on average and a 99.9th percentile of 10 ms by application aware overprovisioning for mission critical traffic and users.

cs.DC

Portability of Scientific Workflows in NGS Data Analysis A Case Study. The analysis of next generation sequencing (NGS) data requires complex computational workflows consisting of dozens of autonomously developed yet interdependent processing steps. Whenever large amounts of data need to be processed, these workflows must be executed on a parallel and or distributed systems to ensure reasonable runtime. Porting a workflow developed for a particular system on a particular hardware infrastructure to another system or to another infrastructure is non trivial, which poses a major impediment to the scientific necessities of workflow reproducibility and workflow reusability. In this work, we describe our efforts to port a state of the art workflow for the detection of specific variants in whole exome sequencing of mice. The workflow originally was developed in the scientific workflow system snakemake for execution on a high performance cluster controlled by Sun Grid Engine. In the project, we ported it to the scientific workflow system SaasFee that can execute workflows on (multi core) stand alone servers or on clusters of arbitrary sizes using the Hadoop. The purpose of this port was that also owners of low cost hardware infrastructures, for which Hadoop was made for, become able to use the workflow. Although both the source and the target system are called scientific workflow systems, they differ in numerous aspects, ranging from the workflow languages to the scheduling mechanisms and the file access interfaces. These differences resulted in various problems, some expected and more unexpected, that had to be resolved before the workflow could be run with equal semantics. As a side effect, we also report cost runtime ratios for a state of the art NGS workflow on very different hardware platforms A comparably cheap stand alone server (80 threads), a mid cost, mid sized cluster (552 threads), and a high end HPC system (3784 threads).

math.DS math.SG

Generalized k contact structures. With the goal to study and better understand algebraic Anosov actions of mathbb R k , we develop a higher codimensional analogue of the contact distribution on odd dimensional manifolds, call such structure a generalized k contact structure. We show that there exist an mathbb R k action associated with this structure, afterwards, we relate this structure with the Weyl chamber actions and a few more general algebraic Anosov actions, proving that such actions admits a compatible generalized k contact structure.

cs.CV

RobustScanner Dynamically Enhancing Positional Clues for Robust Text Recognition. The attention based encoder decoder framework has recently achieved impressive results for scene text recognition, and many variants have emerged with improvements in recognition quality. However, it performs poorly on contextless texts (e.g., random character sequences) which is unacceptable in most of real application scenarios. In this paper, we first deeply investigate the decoding process of the decoder. We empirically find that a representative character level sequence decoder utilizes not only context information but also positional information. Contextual information, which the existing approaches heavily rely on, causes the problem of attention drift. To suppress such side effect, we propose a novel position enhancement branch, and dynamically fuse its outputs with those of the decoder attention module for scene text recognition. Specifically, it contains a position aware module to enable the encoder to output feature vectors encoding their own spatial positions, and an attention module to estimate glimpses using the positional clue (i.e., the current decoding time step) only. The dynamic fusion is conducted for more robust feature via an element wise gate mechanism. Theoretically, our proposed method, dubbed emph RobustScanner , decodes individual characters with dynamic ratio between context and positional clues, and utilizes more positional ones when the decoding sequences with scarce context, and thus is robust and practical. Empirically, it has achieved new state of the art results on popular regular and irregular text recognition benchmarks while without much performance drop on contextless benchmarks, validating its robustness in both contextual and contextless application scenarios.

math.PR math.CA

On the strong regularity of degenerate additive noise driven stochastic differential equations with respect to their initial values. Recently in M. Hairer, M. Hutzenthaler, and A. Jentzen, Ann. Probab. 43, 2 (2015), 468 527 and A. Jentzen, T. M "uller Gronbach, and L. Yaroslavtseva, Commun. Math. Sci. 14, 6 (2016), 1477 1500 stochastic differential equations (SDEs) with smooth coefficient functions have been constructed which have an arbitrarily slowly converging modulus of continuity in the initial value. In these SDEs it is crucial that some of the first order partial derivatives of the drift coefficient functions grow at least exponentially and, in particular, quicker than any polynomial. However, in applications SDEs do typically have coefficient functions whose first order partial derivatives are polynomially bounded. In this article we study whether arbitrarily bad regularity phenomena in the initial value may also arise in the latter case and we partially answer this question in the negative. More precisely, we show that every additive noise driven SDE which admits a Lyapunov type condition (which ensures the existence of a unique solution of the SDE) and which has a drift coefficient function whose first order partial derivatives grow at most polynomially is at least logarithmically H "older continuous in the initial value.

math.AP math.DG

Stability of the non abelian X ray transform in dimension ge 3 . Non abelian X ray tomography seeks to recover a matrix potential Phi M rightarrow mathbb C m times m in a domain M from measurements of its so called scattering data C Phi at partial M . For dim M ge 3 (and under appropriate convexity and regularity conditions), injectivity of the forward map Phi mapsto C Phi was established in arXiv 1605.07894 . In this article we extend arXiv 1605.07894 by proving a H "older type stability estimate. As an application we generalise a statistical consistency result for dim M 2 arXiv 1905.00860 to higher dimensions. The injectivity proof in arXiv 1605.07894 relies on a novel method by Uhlmann Vasy arXiv 1210.2084 , which first establishes injectivity in a shallow layer below partial M and then globalises this by a layer stripping argument. The main technical contribution of this paper is a more quantitative version of these arguments, in particular proving uniform bounds on layer depth and stability constants.

cs.CV

Real Time Quality Assessment of Pediatric MRI via Semi Supervised Deep Nonlocal Residual Neural Networks. In this paper, we introduce an image quality assessment (IQA) method for pediatric T1 and T2 weighted MR images. IQA is first performed slice wise using a nonlocal residual neural network (NR Net) and then volume wise by agglomerating the slice QA results using random forest. Our method requires only a small amount of quality annotated images for training and is designed to be robust to annotation noise that might occur due to rater errors and the inevitable mix of good and bad slices in an image volume. Using a small set of quality assessed images, we pre train NR Net to annotate each image slice with an initial quality rating (i.e., pass, questionable, fail), which we then refine by semi supervised learning and iterative self training. Experimental results demonstrate that our method, trained using only samples of modest size, exhibit great generalizability, capable of real time (milliseconds per volume) large scale IQA with near perfect accuracy.

cs.NI

Energy Minimization for Mobile Edge Computing Networks with Time Sensitive Constraints. Mobile edge computing (MEC) provides users with a high quality experience (QoE) by placing servers with rich services close to the end users. Compared with local computing, MEC can contribute to energy saving, but results in increased communication latency. In this paper, we jointly optimize task offloading and resource allocation to minimize the energy consumption in an orthogonal frequency division multiple access (OFDMA) based MEC networks, where the time sensitive tasks can be processed at both local users and MEC server via partial offloading. Since the optimization variables of the problem are strongly coupled, we first decompose the original problem into two subproblems named as offloading selection (PO), and subcarriers and computing resource allocation (PS), and then propose an iterative algorithm to deal with them in a sequence. To be specific, we derive the closed form solution for PO, and deal with PS by an alternating way in the dual domain due to its NP hardness. Simulation results demonstrate

cs.CR cs.CY

Dynamic real time risk analytics of uncontrollable states in complex internet of things systems, cyber risk at the edge. The Internet of Things (IoT) triggers new types of cyber risks. Therefore, the integration of new IoT devices and services requires a self assessment of IoT cyber security posture. By security posture this article refers to the cybersecurity strength of an organisation to predict, prevent and respond to cyberthreats. At present, there is a gap in the state of the art, because there are no self assessment methods for quantifying IoT cyber risk posture. To address this gap, an empirical analysis is performed of 12 cyber risk assessment approaches. The results and the main findings from the analysis is presented as the current and a target risk state for IoT systems, followed by conclusions and recommendations on a transformation roadmap, describing how IoT systems can achieve the target state with a new goal oriented dependency model. By target state, we refer to the cyber security target that matches the generic security requirements of an organisation. The research paper studies and adapts four alternatives for IoT risk assessment and identifies the goal oriented dependency modelling as a dominant approach among the risk assessment models studied. The new goal oriented dependency model in this article enables the assessment of uncontrollable risk states in complex IoT systems and can be used for a quantitative self assessment of IoT cyber risk posture.

math.CA

Asymptotically best possible Lebesque type inequalities for the Fourier sums on sets of generalized Poisson integrals. In this paper we establish Lebesgue type inequalities for 2 pi periodic functions f , which are defined by generalized Poisson integrals of the functions varphi from L p , 1 leq p infty . In these inequalities uniform norms of deviations of Fourier sums f S n 1 C are expressed via best approximations E n ( varphi) L p of functions varphi by trigonometric polynomials in the metric of space L p . We show that obtained estimates are asymptotically best possible.

cs.CV

SMAP Single Shot Multi Person Absolute 3D Pose Estimation. Recovering multi person 3D poses with absolute scales from a single RGB image is a challenging problem due to the inherent depth and scale ambiguity from a single view. Addressing this ambiguity requires to aggregate various cues over the entire image, such as body sizes, scene layouts, and inter person relationships. However, most previous methods adopt a top down scheme that first performs 2D pose detection and then regresses the 3D pose and scale for each detected person individually, ignoring global contextual cues. In this paper, we propose a novel system that first regresses a set of 2.5D representations of body parts and then reconstructs the 3D absolute poses based on these 2.5D representations with a depth aware part association algorithm. Such a single shot bottom up scheme allows the system to better learn and reason about the inter person depth relationship, improving both 3D and 2D pose estimation. The experiments demonstrate that the proposed approach achieves the state of the art performance on the CMU Panoptic and MuPoTS 3D datasets and is applicable to in the wild videos.

q-bio.PE

Lithic techno complexes in Italy from 50 to 39 thousand years BP an overview of lithic technological changes across the Middle Upper Palaeolithic boundary. Defining the processes involved in the technical cultural shifts from the Late Middle to the Early Upper Palaeolithic in Europe ( 50 39 thousand years BP) is one of the most important tasks facing prehistoric studies. In this debate Italy plays a pivotal role, due to its geographical position between eastern and western Mediterranean Europe as well as to it being the location of several sites showing Late Mousterian, Uluzzian and Protoaurignacian evidence distributed across the Peninsula. Our study aims to provide a synthesis of the available lithic evidence from this key area through a review of the evidence collected from a number of reference sites. The main technical features of the Late Mousterian, the Uluzzian and the Protoaurignacian traditions are examined from a diachronic and spatial perspective. Our overview allows the identification of major differences in the technological behaviour of these populations, making it possible to propose a number of specific working hypotheses on the basis of which further studies can be carried out. This study presents a detailed comparative study of the whole corpus of the lithic production strategies documented during this interval, and crucial element thus emerge.These data are of primary importance in order to assess the nature of the "transition" phenomenon in Italy, thus contributing to the larger debate about the disappearance of Neandertals and the arrival of early Modern Humans in Europe.

physics.optics

Controlling spatiotemporal nonlinearities in multimode fibers with deep neural networks. Spatiotemporal nonlinear interactions in multimode fibers are of interest for beam shaping and frequency conversion by exploiting the nonlinear propagation of different pump regimes from quasi continuous wave to ultrashort pulses centered around visible to infrared pump wavelengths. The nonlinear effects in multi mode fibers depend strongly on the excitation condition, however relatively little work has been reported on this subject. Here, we present the first machine learning approach to learn and control the nonlinear frequency conversion inside multimode fibers by tailoring the excitation condition via deep neural networks. Trained with experimental data, deep neural networks are adapted to learn the relation between the spatial beam profile of the pump pulse and the spectrum generation. For different user defined target spectra, network suggested beam shapes are applied and control over the cascaded Raman scattering and supercontinuum generation processes are achieved. Our results present a novel method to tune the spectra of a broadband source.

math.AP

Asymptotic behaviour of singular solution of the fast diffusion equation in the punctured Euclidean space. For n ge 3 , 0 m frac n 2 n , beta 0 and alpha frac 2 beta 1 m , we prove the existence, uniqueness and asymptotics near the origin of the singular eternal self similar solutions of the fast diffusion equation in ( mathbb R n setminus 0 ) times mathbb R of the form U lambda (x,t) e alpha t f lambda (e beta t x), x in mathbb R n setminus 0 , t in mathbb R , where f lambda is a radially symmetric function satisfying frac n 1 m Delta f m alpha f beta x cdot nabla f 0 text in mathbb R n setminus 0 , with underset substack r to 0 lim frac r 2f(r) 1 m log r 1 frac 2(n 1)(n 2 nm) beta (1 m) and underset substack r to infty lim r frac n 2 m f(r) lambda frac 2 1 m frac n 2 m , for some constant lambda 0 . As a consequence we prove the existence and uniqueness of solutions of Cauchy problem for the fast diffusion equation u t frac n 1 m Delta u m in ( mathbb R n setminus 0 ) times (0, infty) with initial value u 0 satisfying f lambda 1 (x) le u 0(x) le f lambda 2 (x) , forall x in mathbb R n setminus 0 , which satisfies U lambda 1 (x,t) le u(x,t) le U lambda 2 (x,t) , forall x in mathbb R n setminus 0 , t ge 0 , for some constants lambda 1 lambda 2 0 . We also prove the asymptotic behaviour of such singular solution u of the fast diffusion equation as t to infty when n 3,4 and frac n 2 n 2 le m frac n 2 n holds. Asymptotic behaviour of such singular solution u of the fast diffusion equation as t to infty is also obtained when 3 le n 8 , 1 sqrt 2 n le m min left( frac 2(n 2) 3n , frac n 2 n 2 right) , and u(x,t) is radially symmetric in x in mathbb R n setminus 0 for any t 0 under appropriate conditions on the initial value u 0 .

math.NT

Finite and symmetric Mordell Tornheim multiple zeta values. We introduce finite and symmetric Mordell Tornheim type of multiple zeta values and give a new approach to the Kaneko Zagier conjecture stating that the finite and symmetric multiple zeta values satisfy the same relations.

math.OC cs.DC

A Distributed Algorithm for High Dimension Convex Quadratically Constrained Quadratic Programs. We propose a Jacobi style distributed algorithm to solve convex, quadratically constrained quadratic programs (QCQPs), which arise from a broad range of applications. While small to medium sized convex QCQPs can be solved efficiently by interior point algorithms, large scale problems pose significant challenges to traditional algorithms that are mainly designed to be implemented on a single computing unit. The exploding volume of data (and hence, the problem size), however, may overwhelm any such units. In this paper, we propose a distributed algorithm for general, non separable, large scale convex QCQPs, using a novel idea of predictor corrector primal dual update with an adaptive step size. The algorithm enables distributed storage of data as well as parallel distributed computing. We establish the conditions for the proposed algorithm to converge to a global optimum, and implement our algorithm on a computer cluster with multiple nodes using Message Passing Interface (MPI). The numerical experiments are conducted on data sets of various scales from different applications, and the results show that our algorithm exhibits favorable scalability for solving large scale problems.

math.AP

Global Attractors for Hindmarsh Rose Equations in Neurodynamics. Global dynamics of the diffusive and partly diffusive Hindmarsh Rose equations on a three dimensional bounded domain originated in neurodynamics are investigated in this paper. The existence of global attractors as well as the regularity are proved through various uniform estimates showing the dissipative properties and the asymptotically compact characteristics, especially for the partly diffusive Hindmarsh Rose equations by means of the Kolmogorov Riesz theorem.

cs.CV cs.AI

Exploring the Limitations of Behavior Cloning for Autonomous Driving. Driving requires reacting to a wide variety of complex environment conditions and agent behaviors. Explicitly modeling each possible scenario is unrealistic. In contrast, imitation learning can, in theory, leverage data from large fleets of human driven cars. Behavior cloning in particular has been successfully used to learn simple visuomotor policies end to end, but scaling to the full spectrum of driving behaviors remains an unsolved problem. In this paper, we propose a new benchmark to experimentally investigate the scalability and limitations of behavior cloning. We show that behavior cloning leads to state of the art results, including in unseen environments, executing complex lateral and longitudinal maneuvers without these reactions being explicitly programmed. However, we confirm well known limitations (due to dataset bias and overfitting), new generalization issues (due to dynamic objects and the lack of a causal model), and training instability requiring further research before behavior cloning can graduate to real world driving. The code of the studied behavior cloning approaches can be found at https github.com felipecode coiltraine .

cs.RO

Certified Grasping. This paper studies robustness in planar grasping from a geometric perspective. By treating grasping as a process that shapes the free space of an object over time, we can define three types of certificates to guarantee success of a grasp (a) invariance under an initial set, (b) convergence towards a goal grasp, and (c) observability over the final object pose. We develop convex combinatorial models for each of these certificates, which can be expressed as simple semi algebraic relations under mild modeling assumptions. By leveraging these models to synthesize certificates, we optimize certifiable grasps of arbitrary planar objects composed as a union of convex polygons, using manipulators described as point fingers. We validate this approach with simulations and real robot experiments, by grasping random polygons, comparing against other standard grasp planning algorithms, and performing sensorless grasps over different objects.

cs.LG cs.CV

Masksembles for Uncertainty Estimation. Deep neural networks have amply demonstrated their prowess but estimating the reliability of their predictions remains challenging. Deep Ensembles are widely considered as being one of the best methods for generating uncertainty estimates but are very expensive to train and evaluate. MC Dropout is another popular alternative, which is less expensive, but also less reliable. Our central intuition is that there is a continuous spectrum of ensemble like models of which MC Dropout and Deep Ensembles are extreme examples. The first uses an effectively infinite number of highly correlated models while the second relies on a finite number of independent models. To combine the benefits of both, we introduce Masksembles. Instead of randomly dropping parts of the network as in MC dropout, Masksemble relies on a fixed number of binary masks, which are parameterized in a way that allows to change correlations between individual models. Namely, by controlling the overlap between the masks and their density one can choose the optimal configuration for the task at hand. This leads to a simple and easy to implement method with performance on par with Ensembles at a fraction of the cost. We experimentally validate Masksembles on two widely used datasets, CIFAR10 and ImageNet.

math.AG math.CO

Point Line Geometry in the Tropical Plane. We study the classical result by Bruijn and Erd H os regarding the bound on the number of lines determined by a n point configuration in the plane, and in the light of the recently proven Tropical Sylvester Gallai theorem, come up with a tropical version of the above mentioned result. In this work, we introduce stable tropical lines, which help in answering questions pertaining to incidence geometry in the tropical plane. Projective duality in the tropical plane helps in translating the question for stable lines to stable intersections that have been previously studied in depth. Invoking duality between Newton subdivisions and line arrangements, we are able to classify stable intersections with shapes of cells in subdivisions, and this ultimately helps us in coming up with a bound. In this process, we also encounter various unique properties of linear Newton subdivisions which are dual to tropical line arrangements.

cs.RO cs.CV

Knowledge is Never Enough Towards Web Aided Deep Open World Recognition. While today's robots are able to perform sophisticated tasks, they can only act on objects they have been trained to recognize. This is a severe limitation any robot will inevitably see new objects in unconstrained settings, and thus will always have visual knowledge gaps. However, standard visual modules are usually built on a limited set of classes and are based on the strong prior that an object must belong to one of those classes. Identifying whether an instance does not belong to the set of known categories (i.e. open set recognition), only partially tackles this problem, as a truly autonomous agent should be able not only to detect what it does not know, but also to extend dynamically its knowledge about the world. We contribute to this challenge with a deep learning architecture that can dynamically update its known classes in an end to end fashion. The proposed deep network, based on a deep extension of a non parametric model, detects whether a perceived object belongs to the set of categories known by the system and learns it without the need to retrain the whole system from scratch. Annotated images about the new category can be provided by an 'oracle' (i.e. human supervision), or by autonomous mining of the Web. Experiments on two different databases and on a robot platform demonstrate the promise of our approach.

physics.optics

Group velocity distribution and short pulse dispersion in a disordered transverse Anderson localization optical waveguide. We investigate the group velocity distribution of waveguide modes in the presence of disorder. The results are based on extensive numerical simulations of disordered optical waveguides using statistical methods. We observe that the narrowest distribution of group velocities is obtained in the presence of a small amount of disorder therefore, the modal dispersion of an optical pulse is minimized when there is only a slight disorder in the waveguide. The absence of disorder or the presence of a large amount of disorder can result in a large modal dispersion due to the broadening of the distribution of the group velocities. We devise a metric that can be applied to the mode group index probability density function and predict the optimal level of disorder that results in the lowest amount of modal dispersion for short pulse propagation. Our results are important for studying the propagation of optical pulses in the linear regime, e.g., for optical communications and the nonlinear regime for high power short pulse propagation.

math.AG

Gorenstein stable surfaces with K X 2 2 and chi( mathcal O X) 4 . We define and study a concrete stratification of the moduli space of Gorenstein stable surfaces X satisfying K X 2 2 and chi( mathcal O X ) 4 , by first establishing an isomorphism with the moduli space of plane octics with certain singularities, which is then easier to handle concretely. In total, there are 47 inhabited strata with altogether 78 components.

math.DG math.AG

From the signature theorem to anomaly cancellation. We survey the Hirzebruch signature theorem as a special case of the Atiyah Singer index theorem. The family version of the Atiyah Singer index theorem in the form of the Riemann Roch Grothendieck Quillen (RRGQ) formula is then applied to the complexified signature operators varying along the universal family of elliptic curves. The RRGQ formula allows us to determine a generalized cohomology class on the base of the elliptic fibration that is known in physics as (a measure of) the local and global anomaly. Combining several anomalous operators allows us to cancel the local anomaly on a Jacobian elliptic surface, a construction that is based on the construction of the Poincar 'e line bundle over an elliptic surface.

math.PR

Large Deviations and Exit times for reflected McKean Vlasov equations with self stabilizing terms and superlinear drifts. We study a class of reflected McKean Vlasov diffusions over a convex domain with self stabilizing coefficients. This includes coefficients that do not satisfy the classical Wasserstein Lipschitz condition. Further, the process is constrained to a (not necessarily bounded) convex domain by a local time on the boundary. These equations include the subclass of reflected self stabilizing diffusions that drift towards their mean via a convolution of the solution law with a stabilizing potential. Firstly, we establish existence and uniqueness results for this class and address the propagation of chaos. We work with a broad class of coefficients, including drift terms that are locally Lipschitz in spatial and measure variables. However, we do not rely on the boundedness of the domain or the coefficients to account for these non linearities and instead use the self stabilizing properties. We prove a Freidlin Wentzell type Large Deviations Principle and an Eyring Kramer's law for the exit time from subdomains contained in the interior of the reflecting domain.

cs.CV

Improving Panoptic Segmentation at All Scales. Crop based training strategies decouple training resolution from GPU memory consumption, allowing the use of large capacity panoptic segmentation networks on multi megapixel images. Using crops, however, can introduce a bias towards truncating or missing large objects. To address this, we propose a novel crop aware bounding box regression loss (CABB loss), which promotes predictions to be consistent with the visible parts of the cropped objects, while not over penalizing them for extending outside of the crop. We further introduce a novel data sampling and augmentation strategy which improves generalization across scales by counteracting the imbalanced distribution of object sizes. Combining these two contributions with a carefully designed, top down panoptic segmentation architecture, we obtain new state of the art results on the challenging Mapillary Vistas (MVD), Indian Driving and Cityscapes datasets, surpassing the previously best approach on MVD by 4.5 PQ and 5.2 mAP.

q-fin.ST

A New Stock Market Valuation Measure with Applications to Equity Linked Annuities. We propose a new stock market valuation measure. We use the USA stock market data from 1871. We split total returns into three components earnings growth, dividend yield, and valuation change. The first two components are fundamental, the third is speculative. We treat earnings growth as exogenous. Combining the other two components gives us a new valuation measure, which fits autoregression of order 1 with Gaussian innovations, centered at 4.6 . Therefore, long term total returns equal long term earnings growth plus 4.6 . We confirm the classic 4 withdrawal rule for investing in stocks A retiree should invest in stocks and withdraw 4 of initial wealth after adjusting for inflation. This is useful for equity linked annuities, which invest retiree's savings used for annuity purchase into the stock market.

cs.CV cs.LG

Hard Aware Fashion Attribute Classification. Fashion attribute classification is of great importance to many high level tasks such as fashion item search, fashion trend analysis, fashion recommendation, etc. The task is challenging due to the extremely imbalanced data distribution, particularly the attributes with only a few positive samples. In this paper, we introduce a hard aware pipeline to make full use of "hard" samples attributes. We first propose Hard Aware BackPropagation (HABP) to efficiently and adaptively focus on training "hard" data. Then for the identified hard labels, we propose to synthesize more complementary samples for training. To stabilize training, we extend semi supervised GAN by directly deactivating outputs for synthetic complementary samples (Deact). In general, our method is more effective in addressing "hard" cases. HABP weights more on "hard" samples. For "hard" attributes with insufficient training data, Deact brings more stable synthetic samples for training and further improve the performance. Our method is verified on large scale fashion dataset, outperforming other state of the art without any additional supervisions.

cs.CR cs.LG

ATTACK2VEC Leveraging Temporal Word Embeddings to Understand the Evolution of Cyberattacks. Despite the fact that cyberattacks are constantly growing in complexity, the research community still lacks effective tools to easily monitor and understand them. In particular, there is a need for techniques that are able to not only track how prominently certain malicious actions, such as the exploitation of specific vulnerabilities, are exploited in the wild, but also (and more importantly) how these malicious actions factor in as attack steps in more complex cyberattacks. In this paper we present ATTACK2VEC, a system that uses temporal word embeddings to model how attack steps are exploited in the wild, and track how they evolve. We test ATTACK2VEC on a dataset of billions of security events collected from the customers of a commercial Intrusion Prevention System over a period of two years, and show that our approach is effective in monitoring the emergence of new attack strategies in the wild and in flagging which attack steps are often used together by attackers (e.g., vulnerabilities that are frequently exploited together). ATTACK2VEC provides a useful tool for researchers and practitioners to better understand cyberattacks and their evolution, and use this knowledge to improve situational awareness and develop proactive defenses.

math.NA cs.NA

A new interface capturing method for Allen Cahn type equations based on a flow dynamic approach in Lagrangian coordinates, I. One dimensional case. We develop a new Lagrangian approach flow dynamic approach to effectively capture the interface in the Allen Cahn type equations. The underlying principle of this approach is the Energetic Variational Approach (EnVarA), motivated by Rayleigh and Onsager cite onsager1931reciprocal,onsager1931reciprocal2 . Its main advantage, comparing with numerical methods in Eulerian coordinates, is that thin interfaces can be effectively captured with few points in the Lagrangian coordinate. We concentrate in the one dimensional case and construct numerical schemes for the trajectory equation in Lagrangian coordinate that obey the variational structures, and as a consequence, are energy dissipative. Ample numerical results are provided to show that only a fewer points are enough to resolve very thin interfaces by using our Lagrangian approach.

physics.flu-dyn

Linear transient growth of particulate pipe flows. We tackle the question of whether the presence of particles in a pipe flow can influence the linear transient growth of infinitesimal perturbations, in view of better understanding the behaviour of particulate pipe flows in regimes of transition to turbulence. The problem is tackled numerically by means of a simple model where particles are modelled as a second fluid, that interacts with the fluid phase through a two way Stokes drag only. The transient growth is found to be enhanced by the presence of particles, especially so if the particles are localised at a specific radial location of the pipe. At the same time, the mechanisms of transient growth themselves remain those of non particulate flows. The effect is maximised for particles of intermediate size (somewhere between the ballistic limit of very light particles and the point where they are too heavy to be influenced by the flow). Most remarkably, the Segre Silberberg radius (around 2 3 of the pipe radius), where particles naturally cluster in laminar flows, turns out to be close to the optimal location to enhance the transient growth.

math.CA

Nonnegative Whitney Extension Problem for C 1( mathbb R n) . Let f be a real valued function on a compact subset in mathbb R n . We show how to decide if f extends to a nonnegative and C 1 function on mathbb R n . There has been no known result for nonnegative C m extension from a general compact set E when m 0 . The nonnegative extension problem for m geq 2 remains open.

physics.optics physics.app-ph physics.ins-det

Steady information transfer through free space employing orthogonal aberration mode division multiplexing. In this paper we propose and experimentally demonstrate information transfer through free space using a laser beam encoded with multiple orthogonal aberration modes in its phase profile. We use Zernike polynomials which forms a complete set of basis functions to represent the aberration modes. The user information is encoded as amplitudes of multiple Zernike modes whose linear combination is used to define the wavefront of the laser beam in the transmission station. It therefore requires a single phase modulating device since the multiplexing is done digitally providing flexibility over the number and type of modes used. The receiving station uses a wavefront sensor that gives a direct measure of all the aberration modes from a single measurement which is decoded to retrieve the user information. The transmission mechanism enables the use of multiple modes and multiple strengths of each mode to encode the user information and external perturbation compensation, utilizing the completeness of the orthogonal modes.

math.OC cs.NA math.NA

Computing Lyapunov functions using deep neural networks. We propose a deep neural network architecture and a training algorithm for computing approximate Lyapunov functions of systems of nonlinear ordinary differential equations. Under the assumption that the system admits a compositional Lyapunov function, we prove that the number of neurons needed for an approximation of a Lyapunov function with fixed accuracy grows only polynomially in the state dimension, i.e., the proposed approach is able to overcome the curse of dimensionality. We show that nonlinear systems satisfying a small gain condition admit compositional Lyapunov functions. Numerical examples in up to ten space dimensions illustrate the performance of the training scheme.

cs.LG cs.AI cs.CV

Linear Mode Connectivity in Multitask and Continual Learning. Continual (sequential) training and multitask (simultaneous) training are often attempting to solve the same overall objective to find a solution that performs well on all considered tasks. The main difference is in the training regimes, where continual learning can only have access to one task at a time, which for neural networks typically leads to catastrophic forgetting. That is, the solution found for a subsequent task does not perform well on the previous ones anymore. However, the relationship between the different minima that the two training regimes arrive at is not well understood. What sets them apart? Is there a local structure that could explain the difference in performance achieved by the two different schemes? Motivated by recent work showing that different minima of the same task are typically connected by very simple curves of low error, we investigate whether multitask and continual solutions are similarly connected. We empirically find that indeed such connectivity can be reliably achieved and, more interestingly, it can be done by a linear path, conditioned on having the same initialization for both. We thoroughly analyze this observation and discuss its significance for the continual learning process. Furthermore, we exploit this finding to propose an effective algorithm that constrains the sequentially learned minima to behave as the multitask solution. We show that our method outperforms several state of the art continual learning algorithms on various vision benchmarks.

cs.CL

Learning to Ask Unanswerable Questions for Machine Reading Comprehension. Machine reading comprehension with unanswerable questions is a challenging task. In this work, we propose a data augmentation technique by automatically generating relevant unanswerable questions according to an answerable question paired with its corresponding paragraph that contains the answer. We introduce a pair to sequence model for unanswerable question generation, which effectively captures the interactions between the question and the paragraph. We also present a way to construct training data for our question generation models by leveraging the existing reading comprehension dataset. Experimental results show that the pair to sequence model performs consistently better compared with the sequence to sequence baseline. We further use the automatically generated unanswerable questions as a means of data augmentation on the SQuAD 2.0 dataset, yielding 1.9 absolute F1 improvement with BERT base model and 1.7 absolute F1 improvement with BERT large model.

math.CA math.CV

A theory of inner Riesz balayage and its applications. We establish the theory of balayage for the Riesz kernel x y alpha n , alpha in(0,2 , on mathbb R n , n geqslant3 , alternative to that suggested in the book by Landkof. A need for that is caused by the fact that the balayage in that book is defined by means of the integral representation, which, however, so far is not completely justified. Our alternative approach is mainly based on Cartan's ideas concerning inner balayage, formulated by him for the Newtonian kernel. Applying the theory of inner Riesz balayage thereby developed, we obtain a number of criteria for the existence of an inner equilibrium measure gamma A for A subset mathbb R n arbitrary, in particular given in terms of the total mass of the inner swept measure mu A with mu suitably chosen. For example, gamma A exists if and only if varepsilon A ne varepsilon , where varepsilon is a Dirac measure at x 0 and A the inverse of A relative to the sphere x 1 , which leads to a Wiener type criterion of inner alpha irregularity. The results obtained are illustrated by examples.

cs.CL cs.LG

Cold start Active Learning through Self supervised Language Modeling. Active learning strives to reduce annotation costs by choosing the most critical examples to label. Typically, the active learning strategy is contingent on the classification model. For instance, uncertainty sampling depends on poorly calibrated model confidence scores. In the cold start setting, active learning is impractical because of model instability and data scarcity. Fortunately, modern NLP provides an additional source of information pre trained language models. The pre training loss can find examples that surprise the model and should be labeled for efficient fine tuning. Therefore, we treat the language modeling loss as a proxy for classification uncertainty. With BERT, we develop a simple strategy based on the masked language modeling loss that minimizes labeling costs for text classification. Compared to other baselines, our approach reaches higher accuracy within less sampling iterations and computation time.

physics.soc-ph cs.LG math.OC

Optimizing vaccine distribution networks in low and middle income countries. Vaccination has been proven to be the most effective method to prevent infectious diseases. However, there are still millions of children in low and middle income countries who are not covered by routine vaccines and remain at risk. The World Health Organization Expanded Programme on Immunization (WHO EPI) was designed to provide universal childhood vaccine access for children across the world and in this work, we address the design of the distribution network for WHO EPI vaccines. In particular, we formulate the network design problem as a mixed integer program (MIP) and present a new algorithm for typical problems that are too large to be solved using commercial MIP software. We test the algorithm using data derived from four different countries in sub Saharan Africa and show that the algorithm is able to obtain high quality solutions for even the largest problems within a few minutes.

math.CA

Volterra Choquet nonlinear operators. In this paper we study to what extend some properties of the classical linear Volterra operators could be transferred to the nonlinear Volterra Choquet operators, obtained by replacing the classical linear integral with respect to the Lebesgue measure, by the nonlinear Choquet integral with respect to a nonadditive set function. Compactness, Lipschitz and cyclicity properties are studied.

math.OC cs.DM cs.DS math.CO

On Line Sum Optimization. We show that the em column sum optimization problem , of finding a (0,1) matrix with prescribed row sums which minimizes the sum of evaluations of given functions at its column sums, can be solved in polynomial time, either when all functions are the same or when all row sums are bounded by any constant. We conjecture that the more general em line sum optimization problem , of finding a matrix minimizing the sum of given functions evaluated at its row sums and column sums, can also be solved in polynomial time.

math.LO

Prenex normal form theorems in semi classical arithmetic. Akama et al. systematically studied an arithmetical hierarchy of the law of excluded middle and related principles in the context of first order arithmetic. In that paper, they first provide a prenex normal form theorem as a justification of their semi classical principles restricted to prenex formulas. However, there are some errors in their proof. In this paper, we provide a simple counterexample of their prenex normal form theorem, then modify it in an appropriate way. In addition, we characterize several prenex normal form theorems with respect to semi classical arithmetic.

cs.IT math.IT

Optimal Linear Coding Schemes for the Secure Decentralized Pliable Index Coding Problem. We study the secure decentralized Pliable Index CODing (PICOD) problem with circular side information sets at the users. The security constraint forbids every user to decode more than one message while a decentralized setting means there is no central transmitter in the system. Compared to the secure but centralized version of the problem, a converse bound from one of our previous works showed a factor of three difference in optimal code length under the constraint of linear encoding. In this paper, we first list the linearly infeasible cases, that is, problems where no linear code can simultaneously achieve both correctness decodability and security. Then, we propose linear coding schemes for all remaining cases and show that their attained code length is to within an additive constant gap from our converse bound.

cs.CL cs.AI cs.LG

Mitigating Gender Bias for Neural Dialogue Generation with Adversarial Learning. Dialogue systems play an increasingly important role in various aspects of our daily life. It is evident from recent research that dialogue systems trained on human conversation data are biased. In particular, they can produce responses that reflect people's gender prejudice. Many debiasing methods have been developed for various NLP tasks, such as word embedding. However, they are not directly applicable to dialogue systems because they are likely to force dialogue models to generate similar responses for different genders. This greatly degrades the diversity of the generated responses and immensely hurts the performance of the dialogue models. In this paper, we propose a novel adversarial learning framework Debiased Chat to train dialogue models free from gender bias while keeping their performance. Extensive experiments on two real world conversation datasets show that our framework significantly reduces gender bias in dialogue models while maintaining the response quality. The implementation of the proposed framework is released.

physics.flu-dyn cs.LG physics.comp-ph

Deep Learning Convective Flow Using Conditional Generative Adversarial Networks. We developed a general deep learning framework, FluidGAN, that is capable of learning and predicting time dependent convective flow coupled with energy transport. FluidGAN is thoroughly data driven with high speed and accuracy and satisfies the physics of fluid without any prior knowledge of underlying fluid and energy transport physics. FluidGAN also learns the coupling between velocity, pressure and temperature fields. Our framework could be used to learn deterministic multiphysics phenomena where the underlying physical model is complex or unknown.

math.CA

An electrostatic interpretation of the zeros of sieved ultraspherical polynomials. In a companion paper On semiclassical orthogonal polynomials via polynomial mappings, J. Math. Anal. Appl. (2017) we proved that the semiclassical class of orthogonal polynomials is stable under polynomial transformations. In this work we use this fact to derive in an unified way old and new properties concerning the sieved ultraspherical polynomials of the first and second kind. In particular we derive ordinary differential equations for these polynomials. As an application, we use the differential equation for sieved ultraspherical polynomials of the first kind to deduce that the zeros of these polynomials mark the locations of a set of particles that are in electrostatic equilibrium with respect to a particular external field.

cs.CV cs.LG cs.NE

Anomaly Detection in Video Sequence with Appearance Motion Correspondence. Anomaly detection in surveillance videos is currently a challenge because of the diversity of possible events. We propose a deep convolutional neural network (CNN) that addresses this problem by learning a correspondence between common object appearances (e.g. pedestrian, background, tree, etc.) and their associated motions. Our model is designed as a combination of a reconstruction network and an image translation model that share the same encoder. The former sub network determines the most significant structures that appear in video frames and the latter one attempts to associate motion templates to such structures. The training stage is performed using only videos of normal events and the model is then capable to estimate frame level scores for an unknown input. The experiments on 6 benchmark datasets demonstrate the competitive performance of the proposed approach with respect to state of the art methods.