text
null | inputs
dict | prediction
null | prediction_agent
null | annotation
list | annotation_agent
null | multi_label
bool 1
class | explanation
null | id
stringlengths 1
5
| metadata
null | status
stringclasses 2
values | event_timestamp
null | metrics
null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
null | {
"abstract": " We prove a general width duality theorem for combinatorial structures with\nwell-defined notions of cohesion and separation. These might be graphs and\nmatroids, but can be much more general or quite different. The theorem asserts\na duality between the existence of high cohesiveness somewhere local and a\nglobal overall tree structure.\nWe describe cohesive substructures in a unified way in the format of tangles:\nas orientations of low-order separations satisfying certain consistency axioms.\nThese axioms can be expressed without reference to the underlying structure,\nsuch as a graph or matroid, but just in terms of the poset of the separations\nthemselves. This makes it possible to identify tangles, and apply our\ntangle-tree duality theorem, in very diverse settings.\nOur result implies all the classical duality theorems for width parameters in\ngraph minor theory, such as path-width, tree-width, branch-width or rank-width.\nIt yields new, tangle-type, duality theorems for tree-width and path-width. It\nimplies the existence of width parameters dual to cohesive substructures such\nas $k$-blocks, edge-tangles, or given subsets of tangles, for which no width\nduality theorems were previously known.\nAbstract separation systems can be found also in structures quite unlike\ngraphs and matroids. For example, our theorem can be applied to image analysis\nby capturing the regions of an image as tangles of separations defined as\nnatural partitions of its set of pixels. It can be applied in big data contexts\nby capturing clusters as tangles. It can be applied in the social sciences,\ne.g. by capturing as tangles the few typical mindsets of individuals found by a\nsurvey. It could also be applied in pure mathematics, e.g. to separations of\ncompact manifolds.\n",
"title": "Tangle-tree duality in abstract separation systems"
} | null | null | null | null | true | null | 801 | null | Default | null | null |
null | {
"abstract": " Network pruning is aimed at imposing sparsity in a neural network\narchitecture by increasing the portion of zero-valued weights for reducing its\nsize regarding energy-efficiency consideration and increasing evaluation speed.\nIn most of the conducted research efforts, the sparsity is enforced for network\npruning without any attention to the internal network characteristics such as\nunbalanced outputs of the neurons or more specifically the distribution of the\nweights and outputs of the neurons. That may cause severe accuracy drop due to\nuncontrolled sparsity. In this work, we propose an attention mechanism that\nsimultaneously controls the sparsity intensity and supervised network pruning\nby keeping important information bottlenecks of the network to be active. On\nCIFAR-10, the proposed method outperforms the best baseline method by 6% and\nreduced the accuracy drop by 2.6x at the same level of sparsity.\n",
"title": "Attention-Based Guided Structured Sparsity of Deep Neural Networks"
} | null | null | null | null | true | null | 802 | null | Default | null | null |
null | {
"abstract": " Graph models are widely used to analyse diffusion processes embedded in\nsocial contacts and to develop applications. A range of graph models are\navailable to replicate the underlying social structures and dynamics\nrealistically. However, most of the current graph models can only consider\nconcurrent interactions among individuals in the co-located interaction\nnetworks. However, they do not account for indirect interactions that can\ntransmit spreading items to individuals who visit the same locations at\ndifferent times but within a certain time limit. The diffusion phenomena\noccurring through direct and indirect interactions is called same place\ndifferent time (SPDT) diffusion. This paper introduces a model to synthesize\nco-located interaction graphs capturing both direct interactions, where\nindividuals meet at a location, and indirect interactions, where individuals\nvisit the same location at different times within a set timeframe. We analyze\n60 million location updates made by 2 million users from a social networking\napplication to characterize the graph properties, including the space-time\ncorrelations and its time evolving characteristics, such as bursty or ongoing\nbehaviors. The generated synthetic graph reproduces diffusion dynamics of a\nrealistic contact graph, and reduces the prediction error by up to 82% when\ncompare to other contact graph models demonstrating its potential for\nforecasting epidemic spread.\n",
"title": "A Graph Model with Indirect Co-location Links"
} | null | null | [
"Computer Science"
]
| null | true | null | 803 | null | Validated | null | null |
null | {
"abstract": " Training neural networks involves finding minima of a high-dimensional\nnon-convex loss function. Knowledge of the structure of this energy landscape\nis sparse. Relaxing from linear interpolations, we construct continuous paths\nbetween minima of recent neural network architectures on CIFAR10 and CIFAR100.\nSurprisingly, the paths are essentially flat in both the training and test\nlandscapes. This implies that neural networks have enough capacity for\nstructural changes, or that these changes are small between minima. Also, each\nminimum has at least one vanishing Hessian eigenvalue in addition to those\nresulting from trivial invariance.\n",
"title": "Essentially No Barriers in Neural Network Energy Landscape"
} | null | null | null | null | true | null | 804 | null | Default | null | null |
null | {
"abstract": " We study the homogenization process for families of strongly nonlinear\nelliptic systems with the homogeneous Dirichlet boundary conditions. The growth\nand the coercivity of the elliptic operator is assumed to be indicated by a\ngeneral inhomogeneous anisotropic $N-$function, which may be possibly also\ndependent on the spatial variable, i.e., the homogenization process will change\nthe characteristic function spaces at each step. Such a problem is well known\nand there exists many positive results for the function satisfying $\\Delta_2$\nand $\\nabla_2$ conditions an being in addition Hölder continuous with\nrespect to the spatial variable. We shall show that cases these conditions can\nbe neglected and will deal with a rather general problem in general function\nspace setting.\n",
"title": "Homogenization of nonlinear elliptic systems in nonreflexive Musielak-Orlicz spaces"
} | null | null | null | null | true | null | 805 | null | Default | null | null |
null | {
"abstract": " A polynomial $p\\in\\mathbb{R}[z_1,\\dots,z_n]$ is real stable if it has no\nroots in the upper-half complex plane. Gurvits's permanent inequality gives a\nlower bound on the coefficient of the $z_1z_2\\dots z_n$ monomial of a real\nstable polynomial $p$ with nonnegative coefficients. This fundamental\ninequality has been used to attack several counting and optimization problems.\nHere, we study a more general question: Given a stable multilinear polynomial\n$p$ with nonnegative coefficients and a set of monomials $S$, we show that if\nthe polynomial obtained by summing up all monomials in $S$ is real stable, then\nwe can lowerbound the sum of coefficients of monomials of $p$ that are in $S$.\nWe also prove generalizations of this theorem to (real stable) polynomials that\nare not multilinear. We use our theorem to give a new proof of Schrijver's\ninequality on the number of perfect matchings of a regular bipartite graph,\ngeneralize a recent result of Nikolov and Singh, and give deterministic\npolynomial time approximation algorithms for several counting problems.\n",
"title": "A Generalization of Permanent Inequalities and Applications in Counting and Optimization"
} | null | null | null | null | true | null | 806 | null | Default | null | null |
null | {
"abstract": " This volume contains the proceedings of the Fifth International Workshop on\nVerification and Program Transformation (VPT 2017). The workshop took place in\nUppsala, Sweden, on April 29th, 2017, affiliated with the European Joint\nConferences on Theory and Practice of Software (ETAPS). The aim of the VPT\nworkshop series is to provide a forum where people from the areas of program\ntransformation and program verification can fruitfully exchange ideas and gain\na deeper understanding of the interactions between those two fields. Seven\npapers were presented at the workshop. Additionally, three invited talks were\ngiven by Javier Esparza (Technische Universität München, Germany), Manuel\nHermenegildo (IMDEA Software Institute, Madrid, Spain), and Alexey Khoroshilov\n(Linux Verification Center, ISPRAS, Moscow, Russia).\n",
"title": "Proceedings Fifth International Workshop on Verification and Program Transformation"
} | null | null | null | null | true | null | 807 | null | Default | null | null |
null | {
"abstract": " This chapter presents an H-infinity filtering framework for cloud-aided\nsemiactive suspension system with time-varying delays. In this system, road\nprofile information is downloaded from a cloud database to facilitate onboard\nestimation of suspension states. Time-varying data transmission delays are\nconsidered and assumed to be bounded. A quarter-car linear suspension model is\nused and an H-infinity filter is designed with both onboard sensor measurements\nand delayed road profile information from the cloud. The filter design\nprocedure is designed based on linear matrix inequalities (LMIs). Numerical\nsimulation results are reported that illustrates the fusion of cloud-based and\non-board information that can be achieved in Vehicleto- Cloud-to-Vehicle\n(V2C2V) implementation.\n",
"title": "H-infinity Filtering for Cloud-Aided Semi-active Suspension with Delayed Information"
} | null | null | null | null | true | null | 808 | null | Default | null | null |
null | {
"abstract": " In this work we introduce a time- and memory-efficient method for structured\nprediction that couples neuron decisions across both space at time. We show\nthat we are able to perform exact and efficient inference on a densely\nconnected spatio-temporal graph by capitalizing on recent advances on deep\nGaussian Conditional Random Fields (GCRFs). Our method, called VideoGCRF is (a)\nefficient, (b) has a unique global minimum, and (c) can be trained end-to-end\nalongside contemporary deep networks for video understanding. We experiment\nwith multiple connectivity patterns in the temporal domain, and present\nempirical improvements over strong baselines on the tasks of both semantic and\ninstance segmentation of videos.\n",
"title": "Deep Spatio-Temporal Random Fields for Efficient Video Segmentation"
} | null | null | null | null | true | null | 809 | null | Default | null | null |
null | {
"abstract": " This paper introduces a new approach to Large-Eddy Simulation (LES) where\nsubgrid-scale (SGS) dissipation is applied proportionally to the degree of\nlocal spectral broadening, hence mitigated or deactivated in regions dominated\nby large-scale and/or laminar vortical motion. The proposed Coherent vorticity\npreserving (CvP) LES methodology is based on the evaluation of the ratio of the\ntest-filtered to resolved (or grid-filtered) enstrophy $\\sigma$. Values of\n$\\sigma$ close to 1 indicate low sub-test-filter turbulent activity, justifying\nlocal deactivation of the SGS dissipation. The intensity of the SGS dissipation\nis progressively increased for $\\sigma < 1$ which corresponds to a small-scale\nspectral broadening. The SGS dissipation is then fully activated in developed\nturbulence characterized by $\\sigma \\le \\sigma_{eq}$, where the value\n$\\sigma_{eq}$ is derived assuming a Kolmogorov spectrum. The proposed approach\ncan be applied to any eddy-viscosity model, is algorithmically simple and\ncomputationally inexpensive. LES of Taylor-Green vortex breakdown demonstrates\nthat the CvP methodology improves the performance of traditional, non-dynamic\ndissipative SGS models, capturing the peak of total turbulent kinetic energy\ndissipation during transition. Similar accuracy is obtained by adopting\nGermano's dynamic procedure albeit at more than twice the computational\noverhead. A CvP-LES of a pair of unstable periodic helical vortices is shown to\npredict accurately the experimentally observed growth rate using coarse\nresolutions. The ability of the CvP methodology to dynamically sort the\ncoherent, large-scale motion from the smaller, broadband scales during\ntransition is demonstrated via flow visualizations. LES of compressible channel\nare carried out and show a good match with a reference DNS.\n",
"title": "A Coherent vorticity preserving eddy viscosity correction for Large-Eddy Simulation"
} | null | null | null | null | true | null | 810 | null | Default | null | null |
null | {
"abstract": " We consider the problem of inference in a causal generative model where the\nset of available observations differs between data instances. We show how\ncombining samples drawn from the graphical model with an appropriate masking\nfunction makes it possible to train a single neural network to approximate all\nthe corresponding conditional marginal distributions and thus amortize the cost\nof inference. We further demonstrate that the efficiency of importance sampling\nmay be improved by basing proposals on the output of the neural network. We\nalso outline how the same network can be used to generate samples from an\napproximate joint posterior via a chain decomposition of the graph.\n",
"title": "A Universal Marginalizer for Amortized Inference in Generative Models"
} | null | null | null | null | true | null | 811 | null | Default | null | null |
null | {
"abstract": " The weighted Maximum Satisfiability problem (weighted MAX-SAT) is a NP-hard\nproblem with numerous applications arising in artificial intelligence. As an\nefficient tool for heuristic design, the backbone has been applied to\nheuristics design for many NP-hard problems. In this paper, we investigated the\ncomputational complexity for retrieving the backbone in weighted MAX-SAT and\ndeveloped a new algorithm for solving this problem. We showed that it is\nintractable to retrieve the full backbone under the assumption that . Moreover,\nit is intractable to retrieve a fixed fraction of the backbone as well. And\nthen we presented a backbone guided local search (BGLS) with Walksat operator\nfor weighted MAX-SAT. BGLS consists of two phases: the first phase samples the\nbackbone information from local optima and the backbone phase conducts local\nsearch under the guideline of backbone. Extensive experimental results on the\nbenchmark showed that BGLS outperforms the existing heuristics in both solution\nquality and runtime.\n",
"title": "Approximating the Backbone in the Weighted Maximum Satisfiability Problem"
} | null | null | [
"Computer Science"
]
| null | true | null | 812 | null | Validated | null | null |
null | {
"abstract": " We show that in the presence of magnetic field, two superconducting phases\nwith the center-of-mass momentum of Cooper pair parallel to the magnetic field\nare induced in spin-orbit-coupled superconductor Li$_2$Pd$_3$B. Specifically,\nat small magnetic field, the center-of-mass momentum is induced due to the\nenergy-spectrum distortion and no unpairing region with vanishing singlet\ncorrelation appears. We refer to this superconducting state as the drift-BCS\nstate. By further increasing the magnetic field, the superconducting state\nfalls into the Fulde-Ferrell-Larkin-Ovchinnikov state with the emergence of the\nunpairing regions. The observed abrupt enhancement of the center-of-mass\nmomenta and suppression on the order parameters during the crossover indicate\nthe first-order phase transition. Enhanced Pauli limit and hence enlarged\nmagnetic-field regime of the Fulde-Ferrell-Larkin-Ovchinnikov state, due to the\nspin-flip terms of the spin-orbit coupling, are revealed. We also address the\ntriplet correlations induced by the spin-orbit coupling, and show that the\nCooper-pair spin polarizations, generated by the magnetic field and\ncenter-of-mass momentum with the triplet correlations, exhibit totally\ndifferent magnetic-field dependences between the drift-BCS and\nFulde-Ferrell-Larkin-Ovchinnikov states.\n",
"title": "Fulde-Ferrell-Larkin-Ovchinnikov state in spin-orbit-coupled superconductors"
} | null | null | null | null | true | null | 813 | null | Default | null | null |
null | {
"abstract": " We have recently established some integral inequalities for convex functions\nvia the Hermite-Hadamard's inequalities. In continuation here, we also\nestablish some interesting new integral inequalities for convex functions via\nthe Hermite--Hadamard's inequalities and Jensen's integral inequality. Useful\napplications involving special means are also included.\n",
"title": "New type integral inequalities for convex functions with applications II"
} | null | null | null | null | true | null | 814 | null | Default | null | null |
null | {
"abstract": " The support vector machine (SVM) is a powerful and widely used classification\nalgorithm. This paper uses the Karush-Kuhn-Tucker conditions to provide\nrigorous mathematical proof for new insights into the behavior of SVM. These\ninsights provide perhaps unexpected relationships between SVM and two other\nlinear classifiers: the mean difference and the maximal data piling direction.\nFor example, we show that in many cases SVM can be viewed as a cropped version\nof these classifiers. By carefully exploring these connections we show how SVM\ntuning behavior is affected by characteristics including: balanced vs.\nunbalanced classes, low vs. high dimension, separable vs. non-separable data.\nThese results provide further insights into tuning SVM via cross-validation by\nexplaining observed pathological behavior and motivating improved\ncross-validation methodology. Finally, we also provide new results on the\ngeometry of complete data piling directions in high dimensional space.\n",
"title": "Geometric Insights into Support Vector Machine Behavior using the KKT Conditions"
} | null | null | null | null | true | null | 815 | null | Default | null | null |
null | {
"abstract": " Data analytics and data science play a significant role in nowadays society.\nIn the context of Smart Grids (SG), the collection of vast amounts of data has\nseen the emergence of a plethora of data analysis approaches. In this paper, we\nconduct a Systematic Mapping Study (SMS) aimed at getting insights about\ndifferent facets of SG data analysis: application sub-domains (e.g., power load\ncontrol), aspects covered (e.g., forecasting), used techniques (e.g.,\nclustering), tool-support, research methods (e.g., experiments/simulations),\nreplicability/reproducibility of research. The final goal is to provide a view\nof the current status of research. Overall, we found that each sub-domain has\nits peculiarities in terms of techniques, approaches and research methodologies\napplied. Simulations and experiments play a crucial role in many areas. The\nreplicability of studies is limited concerning the provided implemented\nalgorithms, and to a lower extent due to the usage of private datasets.\n",
"title": "Smart Grids Data Analysis: A Systematic Mapping Study"
} | null | null | null | null | true | null | 816 | null | Default | null | null |
null | {
"abstract": " In this paper, we present a new task that investigates how people interact\nwith and make judgments about towers of blocks. In Experiment~1, participants\nin the lab solved a series of problems in which they had to re-configure three\nblocks from an initial to a final configuration. We recorded whether they used\none hand or two hands to do so. In Experiment~2, we asked participants online\nto judge whether they think the person in the lab used one or two hands. The\nresults revealed a close correspondence between participants' actions in the\nlab, and the mental simulations of participants online. To explain\nparticipants' actions and mental simulations, we develop a model that plans\nover a symbolic representation of the situation, executes the plan using a\ngeometric solver, and checks the plan's feasibility by taking into account the\nphysical constraints of the scene. Our model explains participants' actions and\njudgments to a high degree of quantitative accuracy.\n",
"title": "Physical problem solving: Joint planning with symbolic, geometric, and dynamic constraints"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 817 | null | Validated | null | null |
null | {
"abstract": " Recent progress in applying complex network theory to problems faced in\nquantum information and computation has resulted in a beneficial crossover\nbetween two fields. Complex network methods have successfully been used to\ncharacterize quantum walk and transport models, entangled communication\nnetworks, graph-theoretic models of emergent space-time and in detecting\nmesoscale structure in quantum systems. Information physics is setting the\nstage for a theory of complex and networked systems with quantum\ninformation-inspired methods appearing in complex network science, including\ninformation-theoretic distance and correlation measures for network\ncharacterization. Novel quantum induced effects have been predicted in random\ngraphs---where edges represent entangled links---and quantum computer\nalgorithms have recently been proposed to offer super-polynomial enhancement\nfor several network and graph theoretic problems. Here we review the results at\nthe cutting edge, pinpointing the similarities and reconciling the differences\nfound in the series of results at the intersection of these two fields.\n",
"title": "Complex Networks: from Classical to Quantum"
} | null | null | null | null | true | null | 818 | null | Default | null | null |
null | {
"abstract": " Under suitable conditions, a substitution tiling gives rise to a Smale space,\nfrom which three equivalence relations can be constructed, namely the stable,\nunstable, and asymptotic equivalence relations. We denote with $S$, $U$, and\n$A$ their corresponding $C^*$-algebras in the sense of Renault. In this article\nwe show that the $K$-theories of $S$ and $U$ can be computed from the\ncohomology and homology of a single cochain complex with connecting maps for\ntilings of the line and of the plane. Moreover, we provide formulas to compute\nthe $K$-theory for these three $C^*$-algebras. Furthermore, we show that the\n$K$-theory groups for tilings of dimension 1 are always torsion free. For\ntilings of dimension 2, only $K_0(U)$ and $K_1(S)$ can contain torsion.\n",
"title": "On the K-theory of C*-algebras for substitution tilings (a pedestrian version)"
} | null | null | [
"Mathematics"
]
| null | true | null | 819 | null | Validated | null | null |
null | {
"abstract": " We prove that the Tutte embeddings (a.k.a. harmonic/embeddings) of certain\nrandom planar maps converge to $\\gamma$-Liouville quantum gravity\n($\\gamma$-LQG). Specifically, we treat mated-CRT maps, which are discretized\nmatings of correlated continuum random trees, and $\\gamma$ ranges from $0$ to\n$2$ as one varies the correlation parameter. We also show that the associated\nspace-filling path on the embedded map converges to space-filling\nSLE$_{\\kappa}$ for $\\kappa =16/\\gamma^2$ (in the annealed sense) and that\nsimple random walk on the embedded map converges to Brownian motion (in the\nquenched sense). Our arguments also yield analogous statements for the Smith\n(square tiling) embedding of the mated-CRT map.\nThis work constitutes the first proof that a discrete conformal embedding of\na random planar map converges to LQG. Many more such statements have been\nconjectured. Since the mated-CRT map can be viewed as a coarse-grained\napproximation to other random planar maps (the UIPT, tree-weighted maps,\nbipolar-oriented maps, etc.), our results indicate a potential approach for\nproving that embeddings of these maps converge to LQG as well.\nTo prove the main result, we establish several (independently interesting)\ntheorems about LQG surfaces decorated by space-filling SLE. There is a natural\nway to use the SLE curve to divide the plane into `cells' corresponding to\nvertices of the mated-CRT map. We study the law of the shape of the\norigin-containing cell, in particular proving moments for the ratio of its\nsquared diameter to its area. We also give bounds on the degree of the\norigin-containing cell and establish a form of ergodicity for the entire\nconfiguration. Ultimately, we use these properties to show (using a general\ntheorem proved in a separate paper) that random walk on these cells converges\nto a time change of Brownian motion, which in turn leads to the Tutte embedding\nresult.\n",
"title": "The Tutte embedding of the mated-CRT map converges to Liouville quantum gravity"
} | null | null | [
"Mathematics"
]
| null | true | null | 820 | null | Validated | null | null |
null | {
"abstract": " Muroga [M52] showed how to express the Shannon channel capacity of a discrete\nchannel with noise [S49] as an explicit function of the transition\nprobabilities. His method accommodates channels with any finite number of input\nsymbols, any finite number of output symbols and any transition probability\nmatrix. Silverman [S55] carried out Muroga's method in the special case of a\nbinary channel (and went on to analyse \"cascades\" of several such binary\nchannels).\nThis article is a note on the resulting formula for the capacity C(a, c) of a\nsingle binary channel. We aim to clarify some of the arguments and correct a\nsmall error. In service of this aim, we first formulate several of Shannon's\ndefinitions and proofs in terms of discrete measure-theoretic probability\ntheory. We provide an alternate proof to Silverman's, of the feasibility of the\noptimal input distribution for a binary channel. For convenience, we also\nexpress C(a, c) in a single expression explicitly dependent on a and c only,\nwhich Silverman stopped short of doing.\n",
"title": "On the Computation of the Shannon Capacity of a Discrete Channel with Noise"
} | null | null | null | null | true | null | 821 | null | Default | null | null |
null | {
"abstract": " Let $R$ be a two-sided noetherian ring and $M$ be a nilpotent $R$-bimodule,\nwhich is finitely generated on both sides. We study Gorenstein homological\nproperties of the tensor ring $T_R(M)$. Under certain conditions, the ring $R$\nis Gorenstein if and only if so is $T_R(M)$. We characterize Gorenstein\nprojective $T_R(M)$-modules in terms of $R$-modules.\n",
"title": "Gorenstein homological properties of tensor rings"
} | null | null | null | null | true | null | 822 | null | Default | null | null |
null | {
"abstract": " Regression based methods are not performing as well as detection based\nmethods for human pose estimation. A central problem is that the structural\ninformation in the pose is not well exploited in the previous regression\nmethods. In this work, we propose a structure-aware regression approach. It\nadopts a reparameterized pose representation using bones instead of joints. It\nexploits the joint connection structure to define a compositional loss function\nthat encodes the long range interactions in the pose. It is simple, effective,\nand general for both 2D and 3D pose estimation in a unified setting.\nComprehensive evaluation validates the effectiveness of our approach. It\nsignificantly advances the state-of-the-art on Human3.6M and is competitive\nwith state-of-the-art results on MPII.\n",
"title": "Compositional Human Pose Regression"
} | null | null | null | null | true | null | 823 | null | Default | null | null |
null | {
"abstract": " We consider systems with memory represented by stochastic functional\ndifferential equations. Substantially, these are stochastic differential\nequations with coefficients depending on the past history of the process\nitself. Such coefficients are hence defined on a functional space. Models with\nmemory appear in many applications ranging from biology to finance. Here we\nconsider the results of some evaluations based on these models (e.g. the prices\nof some financial products) and the risks connected to the choice of these\nmodels. In particular we focus on the impact of the initial condition on the\nevaluations. This problem is known as the analysis of sensitivity to the\ninitial condition and, in the terminology of finance, it is referred to as the\nDelta. In this work the initial condition is represented by the relevant past\nhistory of the stochastic functional differential equation. This naturally\nleads to the redesign of the definition of Delta. We suggest to define it as a\nfunctional directional derivative, this is a natural choice. For this we study\na representation formula which allows for its computation without requiring\nthat the evaluation functional is differentiable. This feature is particularly\nrelevant for applications. Our formula is achieved by studying an appropriate\nrelationship between Malliavin derivative and functional directional\nderivative. For this we introduce the technique of {\\it randomisation of the\ninitial condition}.\n",
"title": "Stochastic functional differential equations and sensitivity to their initial path"
} | null | null | null | null | true | null | 824 | null | Default | null | null |
null | {
"abstract": " We discuss the concept of inner function in reproducing kernel Hilbert spaces\nwith an orthogonal basis of monomials and examine connections between inner\nfunctions and optimal polynomial approximants to $1/f$, where $f$ is a function\nin the space. We revisit some classical examples from this perspective, and\nshow how a construction of Shapiro and Shields can be modified to produce inner\nfunctions.\n",
"title": "Remarks on Inner Functions and Optimal Approximants"
} | null | null | null | null | true | null | 825 | null | Default | null | null |
null | {
"abstract": " Data center networks are an important infrastructure in various applications\nof modern information technologies. Note that each data center always has a\nfinite lifetime, thus once a data center fails, then it will lose all its\nstorage files and useful information. For this, it is necessary to replicate\nand copy each important file into other data centers such that this file can\nincrease its lifetime of staying in a data center network. In this paper, we\ndescribe a large-scale data center network with a file d-threshold policy,\nwhich is to replicate each important file into at most d-1 other data centers\nsuch that this file can maintain in the data center network under a given level\nof data security in the long-term. To this end, we develop three relevant\nMarkov processes to propose two effective methods for assessing the file\nlifetime and data security. By using the RG-factorizations, we show that the\ntwo methods are used to be able to more effectively evaluate the file lifetime\nof large-scale data center networks. We hope the methodology and results given\nin this paper are applicable in the file lifetime study of more general data\ncenter networks with replication mechanism.\n",
"title": "A Stochastic Model for File Lifetime and Security in Data Center Networks"
} | null | null | [
"Computer Science"
]
| null | true | null | 826 | null | Validated | null | null |
null | {
"abstract": " In the setting of high-dimensional linear regression models, we propose two\nframeworks for constructing pointwise and group confidence sets for penalized\nestimators which incorporate prior knowledge about the organization of the\nnon-zero coefficients. This is done by desparsifying the estimator as in van de\nGeer et al. [18] and van de Geer and Stucky [17], then using an appropriate\nestimator for the precision matrix $\\Theta$. In order to estimate the precision\nmatrix a corresponding structured matrix norm penalty has to be introduced.\nAfter normalization the result is an asymptotic pivot.\nThe asymptotic behavior is studied and simulations are added to study the\ndifferences between the two schemes.\n",
"title": "Asymptotic Confidence Regions for High-dimensional Structured Sparsity"
} | null | null | null | null | true | null | 827 | null | Default | null | null |
null | {
"abstract": " Finding patterns in data and being able to retrieve information from those\npatterns is an important task in Information retrieval. Complex search\nrequirements which are not fulfilled by simple string matching and require\nexploring certain patterns in data demand a better query engine that can\nsupport searching via structured queries. In this article, we built a\nstructured query engine which supports searching data through structured\nqueries on the lines of ElasticSearch. We will show how we achieved real time\nindexing and retrieving of data through a RESTful API and how complex queries\ncan be created and processed using efficient data structures we created for\nstoring the data in structured way. Finally, we will conclude with an example\nof movie recommendation system built on top of this query engine.\n",
"title": "Building a Structured Query Engine"
} | null | null | [
"Computer Science"
]
| null | true | null | 828 | null | Validated | null | null |
null | {
"abstract": " As a natural extension of compressive sensing and the requirement of some\npractical problems, Phaseless Compressed Sensing (PCS) has been introduced and\nstudied recently. Many theoretical results have been obtained for PCS with the\naid of its convex relaxation. Motivated by successful applications of nonconvex\nrelaxed methods for solving compressive sensing, in this paper, we try to\ninvestigate PCS via its nonconvex relaxation. Specifically, we relax PCS in the\nreal context by the corresponding $\\ell_p$-minimization with $p\\in (0,1)$. We\nshow that there exists a constant $p^\\ast\\in (0,1]$ such that for any fixed\n$p\\in(0, p^\\ast)$, every optimal solution to the $\\ell_p$-minimization also\nsolves the concerned problem; and derive an expression of such a constant\n$p^\\ast$ by making use of the known data and the sparsity level of the\nconcerned problem. These provide a theoretical basis for solving this class of\nproblems via the corresponding $\\ell_p$-minimization.\n",
"title": "A Theoretical Perspective of Solving Phaseless Compressed Sensing via Its Nonconvex Relaxation"
} | null | null | null | null | true | null | 829 | null | Default | null | null |
null | {
"abstract": " We present an exhaustive census of Lyman alpha (Ly$\\alpha$) emission in the\ngeneral galaxy population at $3<z<4.6$. We use the Michigan/Magellan Fiber\nSystem (M2FS) spectrograph to study a stellar mass (M$_*$) selected sample of\n625 galaxies homogeneously distributed in the range\n$7.6<\\log{\\mbox{M$_*$/M$_{\\odot}$}}<10.6$. Our sample is selected from the\n3D-HST/CANDELS survey, which provides the complementary data to estimate\nLy$\\alpha$ equivalent widths ($W_{Ly\\alpha}$) and escape fractions ($f_{esc}$)\nfor our galaxies. We find both quantities to anti-correlate with M$_*$,\nstar-formation rate (SFR), UV luminosity, and UV slope ($\\beta$). We then model\nthe $W_{Ly\\alpha}$ distribution as a function of M$_{UV}$ and $\\beta$ using a\nBayesian approach. Based on our model and matching the properties of typical\nLyman break galaxy (LBG) selections, we conclude that the $W_{Ly\\alpha}$\ndistribution in such samples is heavily dependent on the limiting M$_{UV}$ of\nthe survey. Regarding narrowband surveys, we find their $W_{Ly\\alpha}$\nselections to bias samples toward low M$_*$, while their line-flux limitations\npreferentially leave out low-SFR galaxies. We can also use our model to predict\nthe fraction of Ly$\\alpha$-emitting LBGs at $4\\leqslant z\\leqslant 7$. We show\nthat reported drops in the Ly$\\alpha$ fraction at $z\\geqslant6$, usually\nattributed to the rapidly increasing neutral gas fraction of the universe, can\nalso be explained by survey M$_{UV}$ incompleteness. This result does not\ndismiss reionization occurring at $z\\sim7$, but highlights that current data is\nnot inconsistent with this process taking place at $z>7$.\n",
"title": "A Comprehensive Study of Ly$α$ Emission in the High-redshift Galaxy Population"
} | null | null | null | null | true | null | 830 | null | Default | null | null |
null | {
"abstract": " Building on insights of Jovanovic (1982) and subsequent authors, we develop a\ncomprehensive theory of optimal timing of decisions based around continuation\nvalue functions and operators that act on them. Optimality results are provided\nunder general settings, with bounded or unbounded reward functions. This\napproach has several intrinsic advantages that we exploit in developing the\ntheory. One is that continuation value functions are smoother than value\nfunctions, allowing for sharper analysis of optimal policies and more efficient\ncomputation. Another is that, for a range of problems, the continuation value\nfunction exists in a lower dimensional space than the value function,\nmitigating the curse of dimensionality. In one typical experiment, this reduces\nthe computation time from over a week to less than three minutes.\n",
"title": "Optimal Timing of Decisions: A General Theory Based on Continuation Values"
} | null | null | null | null | true | null | 831 | null | Default | null | null |
null | {
"abstract": " OSIRIS-REx will return pristine samples of carbonaceous asteroid Bennu. This\narticle describes how pristine was defined based on expectations of Bennu and\non a realistic understanding of what is achievable with a constrained schedule\nand budget, and how that definition flowed to requirements and implementation.\nTo return a pristine sample, the OSIRIS- REx spacecraft sampling hardware was\nmaintained at level 100 A/2 and <180 ng/cm2 of amino acids and hydrazine on the\nsampler head through precision cleaning, control of materials, and vigilance.\nContamination is further characterized via witness material exposed to the\nspacecraft assembly and testing environment as well as in space. This\ncharacterization provided knowledge of the expected background and will be used\nin conjunction with archived spacecraft components for comparison with the\nsamples when they are delivered to Earth for analysis. Most of all, the\ncleanliness of the OSIRIS-REx spacecraft was achieved through communication\namong scientists, engineers, managers, and technicians.\n",
"title": "OSIRIS-REx Contamination Control Strategy and Implementation"
} | null | null | null | null | true | null | 832 | null | Default | null | null |
null | {
"abstract": " We consider $d$-dimensional linear stochastic approximation algorithms (LSAs)\nwith a constant step-size and the so called Polyak-Ruppert (PR) averaging of\niterates. LSAs are widely applied in machine learning and reinforcement\nlearning (RL), where the aim is to compute an appropriate $\\theta_{*} \\in\n\\mathbb{R}^d$ (that is an optimum or a fixed point) using noisy data and $O(d)$\nupdates per iteration. In this paper, we are motivated by the problem (in RL)\nof policy evaluation from experience replay using the \\emph{temporal\ndifference} (TD) class of learning algorithms that are also LSAs. For LSAs with\na constant step-size, and PR averaging, we provide bounds for the mean squared\nerror (MSE) after $t$ iterations. We assume that data is \\iid with finite\nvariance (underlying distribution being $P$) and that the expected dynamics is\nHurwitz. For a given LSA with PR averaging, and data distribution $P$\nsatisfying the said assumptions, we show that there exists a range of constant\nstep-sizes such that its MSE decays as $O(\\frac{1}{t})$.\nWe examine the conditions under which a constant step-size can be chosen\nuniformly for a class of data distributions $\\mathcal{P}$, and show that not\nall data distributions `admit' such a uniform constant step-size. We also\nsuggest a heuristic step-size tuning algorithm to choose a constant step-size\nof a given LSA for a given data distribution $P$. We compare our results with\nrelated work and also discuss the implication of our results in the context of\nTD algorithms that are LSAs.\n",
"title": "Linear Stochastic Approximation: Constant Step-Size and Iterate Averaging"
} | null | null | null | null | true | null | 833 | null | Default | null | null |
null | {
"abstract": " We study the fundamental tradeoffs between statistical accuracy and\ncomputational tractability in the analysis of high dimensional heterogeneous\ndata. As examples, we study sparse Gaussian mixture model, mixture of sparse\nlinear regressions, and sparse phase retrieval model. For these models, we\nexploit an oracle-based computational model to establish conjecture-free\ncomputationally feasible minimax lower bounds, which quantify the minimum\nsignal strength required for the existence of any algorithm that is both\ncomputationally tractable and statistically accurate. Our analysis shows that\nthere exist significant gaps between computationally feasible minimax risks and\nclassical ones. These gaps quantify the statistical price we must pay to\nachieve computational tractability in the presence of data heterogeneity. Our\nresults cover the problems of detection, estimation, support recovery, and\nclustering, and moreover, resolve several conjectures of Azizyan et al. (2013,\n2015); Verzelen and Arias-Castro (2017); Cai et al. (2016). Interestingly, our\nresults reveal a new but counter-intuitive phenomenon in heterogeneous data\nanalysis that more data might lead to less computation complexity.\n",
"title": "Curse of Heterogeneity: Computational Barriers in Sparse Mixture Models and Phase Retrieval"
} | null | null | null | null | true | null | 834 | null | Default | null | null |
null | {
"abstract": " Audio-visual speech recognition (AVSR) system is thought to be one of the\nmost promising solutions for robust speech recognition, especially in noisy\nenvironment. In this paper, we propose a novel multimodal attention based\nmethod for audio-visual speech recognition which could automatically learn the\nfused representation from both modalities based on their importance. Our method\nis realized using state-of-the-art sequence-to-sequence (Seq2seq)\narchitectures. Experimental results show that relative improvements from 2% up\nto 36% over the auditory modality alone are obtained depending on the different\nsignal-to-noise-ratio (SNR). Compared to the traditional feature concatenation\nmethods, our proposed approach can achieve better recognition performance under\nboth clean and noisy conditions. We believe modality attention based end-to-end\nmethod can be easily generalized to other multimodal tasks with correlated\ninformation.\n",
"title": "Modality Attention for End-to-End Audio-visual Speech Recognition"
} | null | null | null | null | true | null | 835 | null | Default | null | null |
null | {
"abstract": " The goal of this survey article is to explain and elucidate the affine\nstructure of recent models appearing in the rough volatility literature, and\nshow how it leads to exponential-affine transform formulas.\n",
"title": "Affine Rough Models"
} | null | null | [
"Quantitative Finance"
]
| null | true | null | 836 | null | Validated | null | null |
null | {
"abstract": " White dwarf stars have been used as flux standards for decades, thanks to\ntheir staid simplicity. We have empirically tested their photometric stability\nby analyzing the light curves of 398 high-probability candidates and\nspectroscopically confirmed white dwarfs observed during the original Kepler\nmission and later with K2 Campaigns 0-8. We find that the vast majority (>97\nper cent) of non-pulsating and apparently isolated white dwarfs are stable to\nbetter than 1 per cent in the Kepler bandpass on 1-hr to 10-d timescales,\nconfirming that these stellar remnants are useful flux standards. From the\ncases that do exhibit significant variability, we caution that binarity,\nmagnetism, and pulsations are three important attributes to rule out when\nestablishing white dwarfs as flux standards, especially those hotter than\n30,000 K.\n",
"title": "When flux standards go wild: white dwarfs in the age of Kepler"
} | null | null | null | null | true | null | 837 | null | Default | null | null |
null | {
"abstract": " Most existing approaches address multi-view subspace clustering problem by\nconstructing the affinity matrix on each view separately and afterwards propose\nhow to extend spectral clustering algorithm to handle multi-view data. This\npaper presents an approach to multi-view subspace clustering that learns a\njoint subspace representation by constructing affinity matrix shared among all\nviews. Relying on the importance of both low-rank and sparsity constraints in\nthe construction of the affinity matrix, we introduce the objective that\nbalances between the agreement across different views, while at the same time\nencourages sparsity and low-rankness of the solution. Related low-rank and\nsparsity constrained optimization problem is for each view solved using the\nalternating direction method of multipliers. Furthermore, we extend our\napproach to cluster data drawn from nonlinear subspaces by solving the\ncorresponding problem in a reproducing kernel Hilbert space. The proposed\nalgorithm outperforms state-of-the-art multi-view subspace clustering\nalgorithms on one synthetic and four real-world datasets.\n",
"title": "Multi-view Low-rank Sparse Subspace Clustering"
} | null | null | null | null | true | null | 838 | null | Default | null | null |
null | {
"abstract": " The P300 event-related potential (ERP), evoked in scalp-recorded\nelectroencephalography (EEG) by external stimuli, has proven to be a reliable\nresponse for controlling a BCI. The P300 component of an event related\npotential is thus widely used in brain-computer interfaces to translate the\nsubjects' intent by mere thoughts into commands to control artificial devices.\nThe main challenge in the classification of P300 trials in\nelectroencephalographic (EEG) data is the low signal-to-noise ratio (SNR) of\nthe P300 response. To overcome the low SNR of individual trials, it is common\npractice to average together many consecutive trials, which effectively\ndiminishes the random noise. Unfortunately, when more repeated trials are\nrequired for applications such as the P300 speller, the communication rate is\ngreatly reduced. This has resulted in a need for better methods to improve\nsingle-trial classification accuracy of P300 response. In this work, we use\nPrincipal Component Analysis (PCA) as a preprocessing method and use Linear\nDiscriminant Analysis (LDA)and neural networks for classification. The results\nshow that a combination of PCA with these methods provided as high as 13\\%\naccuracy gain for single-trial classification while using only 3 to 4 principal\ncomponents.\n",
"title": "Single-trial P300 Classification using PCA with LDA, QDA and Neural Networks"
} | null | null | null | null | true | null | 839 | null | Default | null | null |
null | {
"abstract": " The purpose of this paper is to study stable representations of partially\nordered sets (posets) and compare it to the well known theory for quivers. In\nparticular, we prove that every indecomposable representation of a poset of\nfinite type is stable with respect to some weight and construct that weight\nexplicitly in terms of the dimension vector. We show that if a poset is\nprimitive then Coxeter transformations preserve stable representations. When\nthe base field is the field of complex numbers we establish the connection\nbetween the polystable representations and the unitary $\\chi$-representations\nof posets. This connection explains the similarity of the results obtained in\nthe series of papers.\n",
"title": "Stable representations of posets"
} | null | null | null | null | true | null | 840 | null | Default | null | null |
null | {
"abstract": " Segmentation in dynamic outdoor environments can be difficult when the\nillumination levels and other aspects of the scene cannot be controlled.\nSpecifically in orchard and vineyard automation contexts, a background material\nis often used to shield a camera's field of view from other rows of crops. In\nthis paper, we describe a method that uses superpixels to determine low texture\nregions of the image that correspond to the background material, and then show\nhow this information can be integrated with the color distribution of the image\nto compute optimal segmentation parameters to segment objects of interest.\nQuantitative and qualitative experiments demonstrate the suitability of this\napproach for dynamic outdoor environments, specifically for tree reconstruction\nand apple flower detection applications.\n",
"title": "Automatic segmentation of trees in dynamic outdoor environments"
} | null | null | [
"Computer Science"
]
| null | true | null | 841 | null | Validated | null | null |
null | {
"abstract": " Stochastic bandit algorithms can be used for challenging non-convex\noptimization problems. Hyperparameter tuning of neural networks is particularly\nchallenging, necessitating new approaches. To this end, we present a method\nthat adaptively partitions the combined space of hyperparameters, context, and\ntraining resources (e.g., total number of training iterations). By adaptively\npartitioning the space, the algorithm is able to focus on the portions of the\nhyperparameter search space that are most relevant in a practical way. By\nincluding the resources in the combined space, the method tends to use fewer\ntraining resources overall. Our experiments show that this method can surpass\nstate-of-the-art methods in tuning neural networks on benchmark datasets. In\nsome cases, our implementations can achieve the same levels of accuracy on\nbenchmark datasets as existing state-of-the-art approaches while saving over\n50% of our computational resources (e.g. time, training iterations).\n",
"title": "A Practical Bandit Method with Advantages in Neural Network Tuning"
} | null | null | null | null | true | null | 842 | null | Default | null | null |
null | {
"abstract": " Dynamic security analysis is an important problem of power systems on\nensuring safe operation and stable power supply even when certain faults occur.\nNo matter such faults are caused by vulnerabilities of system components,\nphysical attacks, or cyber-attacks that are more related to cyber-security,\nthey eventually affect the physical stability of a power system. Examples of\nthe loss of physical stability include the Northeast blackout of 2003 in North\nAmerica and the 2015 system-wide blackout in Ukraine. The nonlinear hybrid\nnature, that is, nonlinear continuous dynamics integrated with discrete\nswitching, and the high degree of freedom property of power system dynamics\nmake it challenging to conduct the dynamic security analysis. In this paper, we\nuse the hybrid automaton model to describe the dynamics of a power system and\nmainly deal with the index-1 differential-algebraic equation models regarding\nthe continuous dynamics in different discrete states. The analysis problem is\nformulated as a reachability problem of the associated hybrid model. A\nsampling-based algorithm is then proposed by integrating modeling and\nrandomized simulation of the hybrid dynamics to search for a feasible execution\nconnecting an initial state of the post-fault system and a target set in the\ndesired operation mode. The proposed method enables the use of existing power\nsystem simulators for the synthesis of discrete switching and control\nstrategies through randomized simulation. The effectiveness and performance of\nthe proposed approach are demonstrated with an application to the dynamic\nsecurity analysis of the New England 39-bus benchmark power system exhibiting\nhybrid dynamics. In addition to evaluating the dynamic security, the proposed\nmethod searches for a feasible strategy to ensure the dynamic security of the\nsystem in face of disruptions.\n",
"title": "Dynamic Security Analysis of Power Systems by a Sampling-Based Algorithm"
} | null | null | [
"Computer Science"
]
| null | true | null | 843 | null | Validated | null | null |
null | {
"abstract": " Second order conic programming (SOCP) has been used to model various\napplications in power systems, such as operation and expansion planning. In\nthis paper, we present a two-stage stochastic mixed integer SOCP (MISOCP) model\nfor the distribution system expansion planning problem that considers\nuncertainty and also captures the nonlinear AC power flow. To avoid costly\ninvestment plans due to some extreme scenarios, we further present a\nchance-constrained variant that could lead to cost-effective solutions. To\naddress the computational challenge, we extend the basic Benders decomposition\nmethod and develop a bilinear variant to compute stochastic and\nchance-constrained MISOCP formulations. A set of numerical experiments is\nperformed to illustrate the performance of our models and computational\nmethods. In particular, results show that our Benders decomposition algorithms\ndrastically outperform a professional MISOCP solver in handling stochastic\nscenarios by orders of magnitude.\n",
"title": "Stochastic and Chance-Constrained Conic Distribution System Expansion Planning Using Bilinear Benders Decomposition"
} | null | null | null | null | true | null | 844 | null | Default | null | null |
null | {
"abstract": " We present the multi-hop extensions of the recently proposed energy-efficient\ntime synchronization scheme for wireless sensor networks, which is based on the\nasynchronous source clock frequency recovery and reversed two-way message\nexchanges. We consider two hierarchical extensions based on packet relaying and\ntime-translating gateways, respectively, and analyze their performance with\nrespect to the number of layers and the delay variations through simulations.\nThe simulation results demonstrate that the time synchronization performance of\nthe packet relaying, which has lower complexity, is close to that of\ntime-translating gateways.\n",
"title": "Multi-Hop Extensions of Energy-Efficient Wireless Sensor Network Time Synchronization"
} | null | null | null | null | true | null | 845 | null | Default | null | null |
null | {
"abstract": " Instructional labs are widely seen as a unique, albeit expensive, way to\nteach scientific content. We measured the effectiveness of introductory lab\ncourses at achieving this educational goal across nine different lab courses at\nthree very different institutions. These institutions and courses encompassed a\nbroad range of student populations and instructional styles. The nine courses\nstudied had two key things in common: the labs aimed to reinforce the content\npresented in lectures, and the labs were optional. By comparing the performance\nof students who did and did not take the labs (with careful normalization for\nselection effects), we found universally and precisely no added value to\nlearning from taking the labs as measured by course exam performance. This work\nshould motivate institutions and departments to reexamine the goals and conduct\nof their lab courses, given their resource-intensive nature. We show why these\nresults make sense when looking at the comparative mental processes of students\ninvolved in research and instructional labs, and offer alternative goals and\ninstructional approaches that would make lab courses more educationally\nvaluable.\n",
"title": "Value added or misattributed? A multi-institution study on the educational benefit of labs for reinforcing physics content"
} | null | null | null | null | true | null | 846 | null | Default | null | null |
null | {
"abstract": " The formation of pattern in biological systems may be modeled by a set of\nreaction-diffusion equations. A diffusion-type coupling operator biologically\nsignificant in neuroscience is a difference of Gaussian functions (Mexican Hat\noperator) used as a spatial-convolution kernel. We are interested in the\ndifference among behaviors of \\emph{stochastic} neural field equations, namely\nspace-time stochastic differential-integral equations, and similar\ndeterministic ones. We explore, quantitatively, how the parameters of our model\nthat measure the shape of the coupling kernel, coupling strength, and aspects\nof the spatially-smoothed space-time noise, control the pattern in the\nresulting evolving random field. We find that a spatial pattern that is damped\nin time in a deterministic system may be sustained and amplified by\nstochasticity, most strikingly at an optimal spatio-temporal noise level. In\naddition, we find that spatially-smoothed noise alone causes pattern formation\neven without spatial coupling.\n",
"title": "Smoothed Noise and Mexican Hat Coupling Produce Pattern in a Stochastic Neural Field"
} | null | null | null | null | true | null | 847 | null | Default | null | null |
null | {
"abstract": " Named-entity recognition (NER) aims at identifying entities of interest in a\ntext. Artificial neural networks (ANNs) have recently been shown to outperform\nexisting NER systems. However, ANNs remain challenging to use for non-expert\nusers. In this paper, we present NeuroNER, an easy-to-use named-entity\nrecognition tool based on ANNs. Users can annotate entities using a graphical\nweb-based user interface (BRAT): the annotations are then used to train an ANN,\nwhich in turn predict entities' locations and categories in new texts. NeuroNER\nmakes this annotation-training-prediction flow smooth and accessible to anyone.\n",
"title": "NeuroNER: an easy-to-use program for named-entity recognition based on neural networks"
} | null | null | null | null | true | null | 848 | null | Default | null | null |
null | {
"abstract": " Blocking objects (blockages) between a transmitter and receiver cause\nwireless communication links to transition from line-of-sight (LOS) to\nnon-line-of-sight (NLOS) propagation, which can greatly reduce the received\npower, particularly at higher frequencies such as millimeter wave (mmWave). We\nconsider a cellular network in which a mobile user attempts to connect to two\nor more base stations (BSs) simultaneously, to increase the probability of at\nleast one LOS link, which is a form of macrodiversity. We develop a framework\nfor determining the LOS probability as a function of the number of BSs, when\ntaking into account the correlation between blockages: for example, a single\nblockage close to the device -- including the user's own body -- could block\nmultiple BSs. We consider the impact of the size of blocking objects on the\nsystem reliability probability and show that macrodiversity gains are higher\nwhen the blocking objects are small. We also show that the BS density must\nscale as the square of the blockage density to maintain a given level of\nreliability.\n",
"title": "Macro diversity in Cellular Networks with Random Blockages"
} | null | null | null | null | true | null | 849 | null | Default | null | null |
null | {
"abstract": " We present an introduction to a novel model of an individual and group\nopinion dynamics, taking into account different ways in which different sources\nof information are filtered due to cognitive biases. The agent based model,\nusing Bayesian updating of the individual belief distribution, is based on the\nrecent psychology work by Dan Kahan. Open nature of the model allows to study\nthe effects of both static and time-dependent biases and information processing\nfilters. In particular, the paper compares the effects of two important\npsychological mechanisms: the confirmation bias and the politically motivated\nreasoning. Depending on the effectiveness of the information filtering (agent\nbias), the agents confronted with an objective information source may either\nreach a consensus based on the truth, or remain divided despite the evidence.\nIn general, the model might provide an understanding into the increasingly\npolarized modern societies, especially as it allows mixing of different types\nof filters: psychological, social, and algorithmic.\n",
"title": "Opinion dynamics model based on cognitive biases"
} | null | null | null | null | true | null | 850 | null | Default | null | null |
null | {
"abstract": " Convolutional Neural Networks (CNNs) are commonly thought to recognise\nobjects by learning increasingly complex representations of object shapes. Some\nrecent studies suggest a more important role of image textures. We here put\nthese conflicting hypotheses to a quantitative test by evaluating CNNs and\nhuman observers on images with a texture-shape cue conflict. We show that\nImageNet-trained CNNs are strongly biased towards recognising textures rather\nthan shapes, which is in stark contrast to human behavioural evidence and\nreveals fundamentally different classification strategies. We then demonstrate\nthat the same standard architecture (ResNet-50) that learns a texture-based\nrepresentation on ImageNet is able to learn a shape-based representation\ninstead when trained on \"Stylized-ImageNet\", a stylized version of ImageNet.\nThis provides a much better fit for human behavioural performance in our\nwell-controlled psychophysical lab setting (nine experiments totalling 48,560\npsychophysical trials across 97 observers) and comes with a number of\nunexpected emergent benefits such as improved object detection performance and\npreviously unseen robustness towards a wide range of image distortions,\nhighlighting advantages of a shape-based representation.\n",
"title": "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness"
} | null | null | null | null | true | null | 851 | null | Default | null | null |
null | {
"abstract": " In this paper, we obtain some formulae for harmonic sums, alternating\nharmonic sums and Stirling number sums by using the method of integral\nrepresentations of series. As applications of these formulae, we give explicit\nformula of several quadratic and cubic Euler sums through zeta values and\nlinear sums. Furthermore, some relationships between harmonic numbers and\nStirling numbers of the first kind are established.\n",
"title": "Explicit evaluation of harmonic sums"
} | null | null | [
"Mathematics"
]
| null | true | null | 852 | null | Validated | null | null |
null | {
"abstract": " In this paper we develop cyclic proof systems for the problem of inclusion\nbetween the least sets of models of mutually recursive predicates, when the\nground constraints in the inductive definitions belong to the quantifier-free\nfragments of (i) First Order Logic with the canonical Herbrand interpretation\nand (ii) Separation Logic, respectively. Inspired by classical\nautomata-theoretic techniques of proving language inclusion between tree\nautomata, we give a small set of inference rules, that are proved to be sound\nand complete, under certain semantic restrictions, involving the set of\nconstraints in the inductive system. Moreover, we investigate the decidability\nand computational complexity of these restrictions for all the logical\nfragments considered and provide a proof search semi-algorithm that becomes a\ndecision procedure for the entailment problem, for those systems that fulfill\nthe restrictions.\n",
"title": "Complete Cyclic Proof Systems for Inductive Entailments"
} | null | null | null | null | true | null | 853 | null | Default | null | null |
null | {
"abstract": " In this paper, we introduce a new model for leveraging unlabeled data to\nimprove generalization performances of image classifiers: a two-branch\nencoder-decoder architecture called HybridNet. The first branch receives\nsupervision signal and is dedicated to the extraction of invariant\nclass-related representations. The second branch is fully unsupervised and\ndedicated to model information discarded by the first branch to reconstruct\ninput data. To further support the expected behavior of our model, we propose\nan original training objective. It favors stability in the discriminative\nbranch and complementarity between the learned representations in the two\nbranches. HybridNet is able to outperform state-of-the-art results on CIFAR-10,\nSVHN and STL-10 in various semi-supervised settings. In addition,\nvisualizations and ablation studies validate our contributions and the behavior\nof the model on both CIFAR-10 and STL-10 datasets.\n",
"title": "HybridNet: Classification and Reconstruction Cooperation for Semi-Supervised Learning"
} | null | null | null | null | true | null | 854 | null | Default | null | null |
null | {
"abstract": " We describe a neural network model that jointly learns distributed\nrepresentations of texts and knowledge base (KB) entities. Given a text in the\nKB, we train our proposed model to predict entities that are relevant to the\ntext. Our model is designed to be generic with the ability to address various\nNLP tasks with ease. We train the model using a large corpus of texts and their\nentity annotations extracted from Wikipedia. We evaluated the model on three\nimportant NLP tasks (i.e., sentence textual similarity, entity linking, and\nfactoid question answering) involving both unsupervised and supervised\nsettings. As a result, we achieved state-of-the-art results on all three of\nthese tasks. Our code and trained models are publicly available for further\nacademic research.\n",
"title": "Learning Distributed Representations of Texts and Entities from Knowledge Base"
} | null | null | null | null | true | null | 855 | null | Default | null | null |
null | {
"abstract": " For $n\\ge5$, it is well known that the moduli space $\\mathfrak{M_{0,\\:n}}$ of\nunordered $n$ points on the Riemann sphere is a quotient space of the Zariski\nopen set $K_n$ of $\\mathbb C^{n-3}$ by an $S_n$ action. The stabilizers of this\n$S_n$ action at certain points of this Zariski open set $K_n$ correspond to the\ngroups fixing the sets of $n$ points on the Riemann sphere. Let $\\alpha$ be a\nsubset of $n$ distinct points on the Riemann sphere. We call the group of all\nlinear fractional transformations leaving $\\alpha$ invariant the stabilizer of\n$\\alpha$, which is finite by observation. For each non-trivial finite subgroup\n$G$ of the group ${\\rm PSL}(2,{\\Bbb C})$ of linear fractional transformations,\nwe give the necessary and sufficient condition for finite subsets of the\nRiemann sphere under which the stabilizers of them are conjugate to $G$. We\nalso prove that there does exist some finite subset of the Riemann sphere whose\nstabilizer coincides with $G$. Next we obtain the irreducible decompositions of\nthe representations of the stabilizers on the tangent spaces at the\nsingularities of $\\mathfrak{M_{0,\\:n}}$. At last, on $\\mathfrak{M_{0,\\:5}}$ and\n$\\mathfrak{M_{0,\\:6}}$, we work out explicitly the singularities and the\nrepresentations of their stabilizers on the tangent spaces at them.\n",
"title": "Moduli Spaces of Unordered $n\\ge5$ Points on the Riemann Sphere and Their Singularities"
} | null | null | [
"Mathematics"
]
| null | true | null | 856 | null | Validated | null | null |
null | {
"abstract": " The radio interferometric positioning system (RIPS) is an accurate node\nlocalization method featuring a novel phase-based ranging process. Multipath is\nthe limiting error source for RIPS in ground-deployed scenarios or indoor\napplications. There are four distinct channels involved in the ranging process\nfor RIPS. Multipath reflections affect both the phase and amplitude of the\nranging signal for each channel. By exploiting untapped amplitude information,\nwe put forward a scheme to estimate each channel's multipath profile, which is\nthen subsequently used to correct corresponding errors in phase measurements.\nSimulations show that such a scheme is very effective in reducing multipath\nphase errors, which are essentially brought down to the level of receiver noise\nunder moderate multipath conditions. It is further demonstrated that ranging\nerrors in RIPS are also greatly reduced via the proposed scheme.\n",
"title": "Multipath Error Correction in Radio Interferometric Positioning Systems"
} | null | null | null | null | true | null | 857 | null | Default | null | null |
null | {
"abstract": " We present an affine analog of the evaluation map for quantum groups. Namely\nwe introduce a surjective homomorphism from the quantum toroidal gl(n) algebra\nto the quantum affine gl(n) algebra completed with respect to the homogeneous\ngrading. We give a brief discussion of evaluation modules.\n",
"title": "An evaluation homomorphism for quantum toroidal gl(n) algebras"
} | null | null | null | null | true | null | 858 | null | Default | null | null |
null | {
"abstract": " In this paper we present a framework for risk-sensitive model predictive\ncontrol (MPC) of linear systems affected by stochastic multiplicative\nuncertainty. Our key innovation is to consider a time-consistent, dynamic risk\nevaluation of the cumulative cost as the objective function to be minimized.\nThis framework is axiomatically justified in terms of time-consistency of risk\nassessments, is amenable to dynamic optimization, and is unifying in the sense\nthat it captures a full range of risk preferences from risk-neutral (i.e.,\nexpectation) to worst case. Within this framework, we propose and analyze an\nonline risk-sensitive MPC algorithm that is provably stabilizing. Furthermore,\nby exploiting the dual representation of time-consistent, dynamic risk\nmeasures, we cast the computation of the MPC control law as a convex\noptimization problem amenable to real-time implementation. Simulation results\nare presented and discussed.\n",
"title": "A Framework for Time-Consistent, Risk-Sensitive Model Predictive Control: Theory and Algorithms"
} | null | null | null | null | true | null | 859 | null | Default | null | null |
null | {
"abstract": " Ground-based astronomical observations may be limited by telluric water vapor\nabsorption, which is highly variable in time and significantly complicates both\nspectroscopy and photometry in the near-infrared (NIR). To achieve the\nsensitivity required to detect Earth-sized exoplanets in the NIR, simultaneous\nmonitoring of precipitable water vapor (PWV) becomes necessary to mitigate the\nimpact of variable telluric lines on radial velocity measurements and transit\nlight curves. To address this issue, we present the Camera for the Automatic\nMonitoring of Atmospheric Lines (CAMAL), a stand-alone, inexpensive six-inch\naperture telescope dedicated to measuring PWV at the Fred Lawrence Whipple\nObservatory on Mount Hopkins. CAMAL utilizes three narrowband NIR filters to\ntrace the amount of atmospheric water vapor affecting simultaneous observations\nwith the MINiature Exoplanet Radial Velocity Array (MINERVA) and MINERVA-Red\ntelescopes. Here we present the current design of CAMAL, discuss our data\nanalysis methods, and show results from 11 nights of PWV measurements taken\nwith CAMAL. For seven nights of data, we have independent PWV measurements\nextracted from high-resolution stellar spectra taken with the Tillinghast\nReflector Echelle Spectrometer (TRES) also located on Mount Hopkins. We use the\nTRES spectra to calibrate the CAMAL absolute PWV scale. Comparisons between\nCAMAL and TRES PWV estimates show excellent agreement, matching to within 1 mm\nover a 10 mm range in PWV. Analysis of CAMAL's photometric precision propagates\nto PWV measurements precise to better than 0.5 mm in dry (PWV < 4 mm)\nconditions. We also find that CAMAL-derived PWVs are highly correlated with\nthose from a GPS-based water vapor monitor located approximately 90 km away at\nKitt Peak National Observatory, with a root mean square PWV difference of 0.8\nmm.\n",
"title": "Monitoring Telluric Absorption with CAMAL"
} | null | null | null | null | true | null | 860 | null | Default | null | null |
null | {
"abstract": " Gossip protocols aim at arriving, by means of point-to-point or group\ncommunications, at a situation in which all the agents know each other secrets.\nRecently a number of authors studied distributed epistemic gossip protocols.\nThese protocols use as guards formulas from a simple epistemic logic, which\nmakes their analysis and verification substantially easier.\nWe study here common knowledge in the context of such a logic. First, we\nanalyze when it can be reduced to iterated knowledge. Then we show that the\nsemantics and truth for formulas without nested common knowledge operator are\ndecidable. This implies that implementability, partial correctness and\ntermination of distributed epistemic gossip protocols that use non-nested\ncommon knowledge operator is decidable, as well. Given that common knowledge is\nequivalent to an infinite conjunction of nested knowledge, these results are\nnon-trivial generalizations of the corresponding decidability results for the\noriginal epistemic logic, established in (Apt & Wojtczak, 2016).\nK. R. Apt & D. Wojtczak (2016): On Decidability of a Logic of Gossips. In\nProc. of JELIA 2016, pp. 18-33, doi:10.1007/ 978-3-319-48758-8_2.\n",
"title": "Common Knowledge in a Logic of Gossips"
} | null | null | null | null | true | null | 861 | null | Default | null | null |
null | {
"abstract": " We examine discrete vortex dynamics in two-dimensional flow through a\nnetwork-theoretic approach. The interaction of the vortices is represented with\na graph, which allows the use of network-theoretic approaches to identify key\nvortex-to-vortex interactions. We employ sparsification techniques on these\ngraph representations based on spectral theory for constructing sparsified\nmodels and evaluating the dynamics of vortices in the sparsified setup.\nIdentification of vortex structures based on graph sparsification and sparse\nvortex dynamics are illustrated through an example of point-vortex clusters\ninteracting amongst themselves. We also evaluate the performance of\nsparsification with increasing number of point vortices. The\nsparsified-dynamics model developed with spectral graph theory requires reduced\nnumber of vortex-to-vortex interactions but agrees well with the full nonlinear\ndynamics. Furthermore, the sparsified model derived from the sparse graphs\nconserves the invariants of discrete vortex dynamics. We highlight the\nsimilarities and differences between the present sparsified-dynamics model and\nthe reduced-order models.\n",
"title": "Network-theoretic approach to sparsified discrete vortex dynamics"
} | null | null | null | null | true | null | 862 | null | Default | null | null |
null | {
"abstract": " In this paper we study a non-linear partial differential equation (PDE),\nproposed by N. Kudryashov [arXiv:1611.06813v1[nlin.SI]], using continuum limit\napproximation of mixed Fermi-Pasta-Ulam and Frenkel-Kontorova Models. This\ngeneralized semi-discrete equation can be considered as a model for the\ndescription of non-linear dislocation waves in crystal lattice and the\ncorresponding continuous system can be called mixed generalized potential KdV\nand sine-Gordon equation. We obtain the Bäcklund transformation of this\nequation in Riccati form in inverse method. We further study the\nquasi-integrable deformation of this model.\n",
"title": "Bäcklund Transformation and Quasi-Integrable Deformation of Mixed Fermi-Pasta-Ulam and Frenkel-Kontorova Models"
} | null | null | null | null | true | null | 863 | null | Default | null | null |
null | {
"abstract": " An important, yet largely unstudied, problem in student data analysis is to\ndetect misconceptions from students' responses to open-response questions.\nMisconception detection enables instructors to deliver more targeted feedback\non the misconceptions exhibited by many students in their class, thus improving\nthe quality of instruction. In this paper, we propose a new natural language\nprocessing-based framework to detect the common misconceptions among students'\ntextual responses to short-answer questions. We propose a probabilistic model\nfor students' textual responses involving misconceptions and experimentally\nvalidate it on a real-world student-response dataset. Experimental results show\nthat our proposed framework excels at classifying whether a response exhibits\none or more misconceptions. More importantly, it can also automatically detect\nthe common misconceptions exhibited across responses from multiple students to\nmultiple questions; this property is especially important at large scale, since\ninstructors will no longer need to manually specify all possible misconceptions\nthat students might exhibit.\n",
"title": "Data-Mining Textual Responses to Uncover Misconception Patterns"
} | null | null | null | null | true | null | 864 | null | Default | null | null |
null | {
"abstract": " A Schottky structure on a handlebody $M$ of genus $g$ is provided by a\nSchottky group of rank $g$. A symmetry (an orientation-reversing involution) of\n$M$ is known to have at most $(g+1)$ connected components of fixed points. Each\nof these components is either a point or a compact bordered surface (either\norientable or not) whose boundary is contained in the border of $M$. In this\npaper, we derive sharp upper bounds for the total number of connected\ncomponents of the sets of fixed points of given two or three symmetries of $M$.\nIn order to obtain such an upper bound, we obtain a geometrical structure\ndescription of those extended Kleinian groups $K$ containing a Schottky group\n$\\Gamma$ as finite index normal subgroup so that $K/\\Gamma$ is a dihedral group\n(called dihedral Schottky groups). Our upper bounds turn out to be different to\nthe corresponding ones at the level of closed Riemann surfaces. In contrast to\nthe case of Riemann surfaces, we observe that $M$ cannot have two different\nmaximal symmetries.\n",
"title": "Symmetries of handlebodies and their fixed points: Dihedral extended Schottky groups"
} | null | null | null | null | true | null | 865 | null | Default | null | null |
null | {
"abstract": " In this paper we propose a new method of speaker diarization that employs a\ndeep learning architecture to learn speaker embeddings. In contrast to the\ntraditional approaches that build their speaker embeddings using manually\nhand-crafted spectral features, we propose to train for this purpose a\nrecurrent convolutional neural network applied directly on magnitude\nspectrograms. To compare our approach with the state of the art, we collect and\nrelease for the public an additional dataset of over 6 hours of fully annotated\nbroadcast material. The results of our evaluation on the new dataset and three\nother benchmark datasets show that our proposed method significantly\noutperforms the competitors and reduces diarization error rate by a large\nmargin of over 30% with respect to the baseline.\n",
"title": "Speaker Diarization using Deep Recurrent Convolutional Neural Networks for Speaker Embeddings"
} | null | null | [
"Computer Science"
]
| null | true | null | 866 | null | Validated | null | null |
null | {
"abstract": " This paper describes an implementation of the L-BFGS method designed to deal\nwith two adversarial situations. The first occurs in distributed computing\nenvironments where some of the computational nodes devoted to the evaluation of\nthe function and gradient are unable to return results on time. A similar\nchallenge occurs in a multi-batch approach in which the data points used to\ncompute function and gradients are purposely changed at each iteration to\naccelerate the learning process. Difficulties arise because L-BFGS employs\ngradient differences to update the Hessian approximations, and when these\ngradients are computed using different data points the updating process can be\nunstable. This paper shows how to perform stable quasi-Newton updating in the\nmulti-batch setting, studies the convergence properties for both convex and\nnonconvex functions, and illustrates the behavior of the algorithm in a\ndistributed computing platform on binary classification logistic regression and\nneural network training problems that arise in machine learning.\n",
"title": "A Robust Multi-Batch L-BFGS Method for Machine Learning"
} | null | null | null | null | true | null | 867 | null | Default | null | null |
null | {
"abstract": " Networks of vertically c-oriented prism shaped InN nanowalls, are grown on\nc-GaN/sapphire templates using a CVD technique, where pure indium and ammonia\nare used as metal and nitrogen precursors. A systematic study of the growth,\nstructural and electronic properties of these samples shows a preferential\ngrowth of the islands along [11-20] and [0001] directions leading to the\nformation of such a network structure, where the vertically [0001] oriented\ntapered walls are laterally align along one of the three [11-20] directions.\nInclined facets of these walls are identified as r-planes [(1-102)-planes] of\nwurtzite InN. Onset of absorption for these samples is observed to be higher\nthan the band gap of InN suggesting a high background carrier concentration in\nthis material. Study of the valence band edge through XPS indicates the\nformation of positive depletion regions below the r-plane side facets of the\nwalls. This is in contrast with the observation for c-plane InN epilayers,\nwhere electron accumulation is often reported below the top surface.\n",
"title": "Network of vertically c-oriented prism shaped InN nanowalls grown on c-GaN/sapphire template by chemical vapor deposition technique"
} | null | null | null | null | true | null | 868 | null | Default | null | null |
null | {
"abstract": " Machine Learning focuses on the construction and study of systems that can\nlearn from data. This is connected with the classification problem, which\nusually is what Machine Learning algorithms are designed to solve. When a\nmachine learning method is used by people with no special expertise in machine\nlearning, it is important that the method be robust in classification, in the\nsense that reasonable performance is obtained with minimal tuning of the\nproblem at hand. Algorithms are evaluated based on how robust they can classify\nthe given data. In this paper, we propose a quantifiable measure of robustness,\nand describe a particular learning method that is robust according to this\nmeasure in the context of classification problem. We proposed Adaptive Boosting\n(AdaBoostM1) with J48(C4.5 tree) as a base learner with tuning weight threshold\n(P) and number of iterations (I) for boosting algorithm. To benchmark the\nperformance, we used the baseline classifier, AdaBoostM1 with Decision Stump as\nbase learner without tuning parameters. By tuning parameters and using J48 as\nbase learner, we are able to reduce the overall average error rate ratio\n(errorC/errorNB) from 2.4 to 0.9 for development sets of data and 2.1 to 1.2\nfor evaluation sets of data.\n",
"title": "Enhanced version of AdaBoostM1 with J48 Tree learning method"
} | null | null | null | null | true | null | 869 | null | Default | null | null |
null | {
"abstract": " This paper presents a simple agent-based model of an economic system,\npopulated by agents playing different games according to their different view\nabout social cohesion and tax payment. After a first set of simulations,\ncorrectly replicating results of existing literature, a wider analysis is\npresented in order to study the effects of a dynamic-adaptation rule, in which\ncitizens may possibly decide to modify their individual tax compliance\naccording to individual criteria, such as, the strength of their ethical\ncommitment, the satisfaction gained by consumption of the public good and the\nperceived opinion of neighbors. Results show the presence of thresholds levels\nin the composition of society - between taxpayers and evaders - which explain\nthe extent of damages deriving from tax evasion.\n",
"title": "A Game of Tax Evasion: evidences from an agent-based model"
} | null | null | null | null | true | null | 870 | null | Default | null | null |
null | {
"abstract": " The variability response function (VRF) is generalized to statically\ndeterminate Euler Bernoulli beams with arbitrary stress-strain laws following\nCauchy elastic behavior. The VRF is a Green's function that maps the spectral\ndensity function (SDF) of a statistically homogeneous random field describing\nthe correlation structure of input uncertainty to the variance of a response\nquantity. The appeal of such Green's functions is that the variance can be\ndetermined for any correlation structure by a trivial computation of a\nconvolution integral. The method introduced in this work derives VRFs in closed\nform for arbitrary nonlinear Cauchy-elastic constitutive laws and is\ndemonstrated through three examples. It is shown why and how higher order\nspectra of the random field affect the response variance for nonlinear\nconstitutive laws. In the general sense, the VRF for a statically determinate\nbeam is found to be a matrix kernel whose inner product by a matrix of higher\norder SDFs and statistical moments is integrated to give the response variance.\nThe resulting VRF matrix is unique regardless of the random field's marginal\nprobability density function (PDF) and SDFs.\n",
"title": "Variability response functions for statically determinate beams with arbitrary nonlinear constitutive laws"
} | null | null | null | null | true | null | 871 | null | Default | null | null |
null | {
"abstract": " We show that training a deep network using batch normalization is equivalent\nto approximate inference in Bayesian models. We further demonstrate that this\nfinding allows us to make meaningful estimates of the model uncertainty using\nconventional architectures, without modifications to the network or the\ntraining procedure. Our approach is thoroughly validated by measuring the\nquality of uncertainty in a series of empirical experiments on different tasks.\nIt outperforms baselines with strong statistical significance, and displays\ncompetitive performance with recent Bayesian approaches.\n",
"title": "Bayesian Uncertainty Estimation for Batch Normalized Deep Networks"
} | null | null | null | null | true | null | 872 | null | Default | null | null |
null | {
"abstract": " In this paper we consider a single-cell downlink scenario where a\nmultiple-antenna base station delivers contents to multiple cache-enabled user\nterminals. Based on the multicasting opportunities provided by the so-called\nCoded Caching technique, we investigate three delivery approaches. Our baseline\nscheme employs the coded caching technique on top of max-min fair multicasting.\nThe second one consists of a joint design of Zero-Forcing (ZF) and coded\ncaching, where the coded chunks are formed in the signal domain (complex\nfield). The third scheme is similar to the second one with the difference that\nthe coded chunks are formed in the data domain (finite field). We derive\nclosed-form rate expressions where our results suggest that the latter two\nschemes surpass the first one in terms of Degrees of Freedom (DoF). However, at\nthe intermediate SNR regime forming coded chunks in the signal domain results\nin power loss, and will deteriorate throughput of the second scheme. The main\nmessage of our paper is that the schemes performing well in terms of DoF may\nnot be directly appropriate for intermediate SNR regimes, and modified schemes\nshould be employed.\n",
"title": "Multi-Antenna Coded Caching"
} | null | null | null | null | true | null | 873 | null | Default | null | null |
null | {
"abstract": " The block bootstrap approximates sampling distributions from dependent data\nby resampling data blocks. A fundamental problem is establishing its\nconsistency for the distribution of a sample mean, as a prototypical statistic.\nWe use a structural relationship with subsampling to characterize the bootstrap\nin a new and general manner. While subsampling and block bootstrap differ, the\nblock bootstrap distribution of a sample mean equals that of a $k$-fold\nself-convolution of a subsampling distribution. Motivated by this, we provide\nsimple necessary and sufficient conditions for a convolved subsampling\nestimator to produce a normal limit that matches the target of bootstrap\nestimation. These conditions may be linked to consistency properties of an\noriginal subsampling distribution, which are often obtainable under minimal\nassumptions. Through several examples, the results are shown to validate the\nblock bootstrap for means under significantly weakened assumptions in many\nexisting (and some new) dependence settings, which also addresses a standing\nconjecture of Politis, Romano and Wolf(1999). Beyond sample means, the\nconvolved subsampling estimator may not match the block bootstrap, but instead\nprovides a hybrid-resampling estimator of interest in its own right. For\ngeneral statistics with normal limits, results also establish the consistency\nof convolved subsampling under minimal dependence conditions, including\nnon-stationarity.\n",
"title": "Convolved subsampling estimation with applications to block bootstrap"
} | null | null | null | null | true | null | 874 | null | Default | null | null |
null | {
"abstract": " Given a polynomial system f associated with a simple multiple zero x of\nmultiplicity {\\mu}, we give a computable lower bound on the minimal distance\nbetween the simple multiple zero x and other zeros of f. If x is only given\nwith limited accuracy, we propose a numerical criterion that f is certified to\nhave {\\mu} zeros (counting multiplicities) in a small ball around x.\nFurthermore, for simple double zeros and simple triple zeros whose Jacobian is\nof normalized form, we define modified Newton iterations and prove the\nquantified quadratic convergence when the starting point is close to the exact\nsimple multiple zero. For simple multiple zeros of arbitrary multiplicity whose\nJacobian matrix may not have a normalized form, we perform unitary\ntransformations and modified Newton iterations, and prove its non-quantified\nquadratic convergence and its quantified convergence for simple triple zeros.\n",
"title": "Computing Simple Multiple Zeros of Polynomial Systems"
} | null | null | null | null | true | null | 875 | null | Default | null | null |
null | {
"abstract": " It can be difficult to tell whether a trained generative model has learned to\ngenerate novel examples or has simply memorized a specific set of outputs. In\npublished work, it is common to attempt to address this visually, for example\nby displaying a generated example and its nearest neighbor(s) in the training\nset (in, for example, the L2 metric). As any generative model induces a\nprobability density on its output domain, we propose studying this density\ndirectly. We first study the geometry of the latent representation and\ngenerator, relate this to the output density, and then develop techniques to\ncompute and inspect the output density. As an application, we demonstrate that\n\"memorization\" tends to a density made of delta functions concentrated on the\nmemorized examples. We note that without first understanding the geometry, the\nmeasurement would be essentially impossible to make.\n",
"title": "Latent Geometry and Memorization in Generative Models"
} | null | null | null | null | true | null | 876 | null | Default | null | null |
null | {
"abstract": " Refraction represents one of the most fundamental operations that may be\nperformed by a metasurface. However, simple phasegradient metasurface designs\nsuffer from restricted angular deflection due to spurious diffraction orders.\nIt has been recently shown, using a circuit-based approach, that refraction\nwithout spurious diffraction, or diffraction-free, can fortunately be achieved\nby a transverse metasurface exhibiting either loss-gain or bianisotropy. Here,\nwe rederive these conditions using a medium-based - and hence more insightfull\n- approach based on Generalized Sheet Transition Conditions (GSTCs) and surface\nsusceptibility tensors, and experimentally demonstrate two diffraction-free\nrefractive metasurfaces that are essentially lossless, passive, bianisotropic\nand reciprocal.\n",
"title": "Refracting Metasurfaces without Spurious Diffraction"
} | null | null | null | null | true | null | 877 | null | Default | null | null |
null | {
"abstract": " We reconsider the classic problem of estimating accurately a 2D\ntransformation from point matches between images containing outliers. RANSAC\ndiscriminates outliers by randomly generating minimalistic sampled hypotheses\nand verifying their consensus over the input data. Its response is based on the\nsingle hypothesis that obtained the largest inlier support. In this article we\nshow that the resulting accuracy can be improved by aggregating all generated\nhypotheses. This yields RANSAAC, a framework that improves systematically over\nRANSAC and its state-of-the-art variants by statistically aggregating\nhypotheses. To this end, we introduce a simple strategy that allows to rapidly\naverage 2D transformations, leading to an almost negligible extra computational\ncost. We give practical applications on projective transforms and\nhomography+distortion models and demonstrate a significant performance gain in\nboth cases.\n",
"title": "Accurate Motion Estimation through Random Sample Aggregated Consensus"
} | null | null | [
"Computer Science"
]
| null | true | null | 878 | null | Validated | null | null |
null | {
"abstract": " The Landau collision integral is an accurate model for the small-angle\ndominated Coulomb collisions in fusion plasmas. We investigate a high order\naccurate, fully conservative, finite element discretization of the nonlinear\nmulti-species Landau integral with adaptive mesh refinement using the PETSc\nlibrary (www.mcs.anl.gov/petsc). We develop algorithms and techniques to\nefficiently utilize emerging architectures with an approach that minimizes\nmemory usage and movement and is suitable for vector processing. The Landau\ncollision integral is vectorized with Intel AVX-512 intrinsics and the solver\nsustains as much as 22% of the theoretical peak flop rate of the Second\nGeneration Intel Xeon Phi, Knights Landing, processor.\n",
"title": "Landau Collision Integral Solver with Adaptive Mesh Refinement on Emerging Architectures"
} | null | null | null | null | true | null | 879 | null | Default | null | null |
null | {
"abstract": " In this paper we combine a survey of the most important topological\nproperties of kinematic maps that appear in robotics, with the exposition of\nsome basic results regarding the topological complexity of a map. In\nparticular, we discuss mechanical devices that consist of rigid parts connected\nby joints and show how the geometry of the joints determines the forward\nkinematic map that relates the configuration of joints with the pose of the\nend-effector of the device. We explain how to compute the dimension of the\njoint space and describe topological obstructions for a kinematic map to be a\nfibration or to admit a continuous section. In the second part of the paper we\ndefine the complexity of a continuous map and show how the concept can be\nviewed as a measure of the difficulty to find a robust manipulation plan for a\ngiven mechanical device. We also derive some basic estimates for the complexity\nand relate it to the degree of instability of a manipulation plan.\n",
"title": "A Topologist's View of Kinematic Maps and Manipulation Complexity"
} | null | null | null | null | true | null | 880 | null | Default | null | null |
null | {
"abstract": " The Butler-Portugal algorithm for obtaining the canonical form of a tensor\nexpression with respect to slot symmetries and dummy-index renaming suffers, in\ncertain cases with a high degree of symmetry, from $O(n!)$ explosion in both\ncomputation time and memory. We present a modified algorithm which alleviates\nthis problem in the most common cases---tensor expressions with subsets of\nindices which are totally symmetric or totally antisymmetric---in polynomial\ntime. We also present an implementation of the label-renaming mechanism which\nimproves upon that of the original Butler-Portugal algorithm, thus providing a\nsignificant speed increase for the average case as well as the highly-symmetric\nspecial case. The worst-case behavior remains $O(n!)$, although it occurs in\nmore limited situations unlikely to appear in actual computations. We comment\non possible strategies to take if the nature of a computation should make these\nsituations more likely.\n",
"title": "Faster Tensor Canonicalization"
} | null | null | null | null | true | null | 881 | null | Default | null | null |
null | {
"abstract": " The mass-preconditioning (MP) technique has become a standard tool to enhance\nthe efficiency of the hybrid Monte-Carlo simulation (HMC) of lattice QCD with\ndynamical quarks, for 2-flavors QCD with degenerate quark masses, as well as\nits extension to the case of one-flavor by taking the square-root of the\nfermion determinant of 2-flavors with degenerate masses. However, for lattice\nQCD with domain-wall fermion, the fermion determinant of any single fermion\nflavor can be expressed as a functional integral with an exact pseudofermion\naction $ \\phi^\\dagger H^{-1} \\phi $, where $ H^{-1} $ is a positive-definite\nHermitian operator without taking square-root, and with the chiral structure\n\\cite{Chen:2014hyy}. Consequently, the mass-preconditioning for the exact\none-flavor action (EOFA) does not necessarily follow the conventional (old) MP\npattern. In this paper, we present a new mass-preconditioning for the EOFA,\nwhich is more efficient than the old MP which we have used in Refs.\n\\cite{Chen:2014hyy,Chen:2014bbc}. We perform numerical tests in lattice QCD\nwith $ N_f = 1 $ and $ N_f = 1+1+1+1 $ optimal domain-wall quarks, with one\nmass-preconditioner applied to one of the exact one-flavor actions, and we find\nthat the efficiency of the new MP is more than 20\\% higher than that of the old\nMP.\n",
"title": "Mass Preconditioning for the Exact One-Flavor Action in Lattice QCD with Domain-Wall Fermion"
} | null | null | null | null | true | null | 882 | null | Default | null | null |
null | {
"abstract": " A model in which a three-dimensional elastic medium is represented by a\nnetwork of identical masses connected by springs of random strengths and\nallowed to vibrate only along a selected axis of the reference frame, exhibits\nan Anderson localization transition. To study this transition, we assume that\nthe dynamical matrix of the network is given by a product of a sparse random\nmatrix with real, independent, Gaussian-distributed non-zero entries and its\ntranspose. A finite-time scaling analysis of system's response to an initial\nexcitation allows us to estimate the critical parameters of the localization\ntransition. The critical exponent is found to be $\\nu = 1.57 \\pm 0.02$ in\nagreement with previous studies of Anderson transition belonging to the\nthree-dimensional orthogonal universality class.\n",
"title": "Finite-time scaling at the Anderson transition for vibrations in solids"
} | null | null | null | null | true | null | 883 | null | Default | null | null |
null | {
"abstract": " We analyze the response of a type II superconducting wire to an external\nmagnetic field parallel to it in the framework of Ginzburg-Landau theory. We\nfocus on the surface superconductivity regime of applied field between the\nsecond and third critical values, where the superconducting state survives only\nclose to the sample's boundary. Our first finding is that, in first\napproximation, the shape of the boundary plays no role in determining the\ndensity of superconducting electrons. A second order term is however isolated,\ndirectly proportional to the mean curvature of the boundary. This demonstrates\nthat points of higher boundary curvature (counted inwards) attract\nsuperconducting electrons.\n",
"title": "Universal and shape dependent features of surface superconductivity"
} | null | null | null | null | true | null | 884 | null | Default | null | null |
null | {
"abstract": " Classical principal component analysis (PCA) is not robust to the presence of\nsparse outliers in the data. The use of the $\\ell_1$ norm in the Robust PCA\n(RPCA) method successfully eliminates the weakness of PCA in separating the\nsparse outliers. In this paper, by sticking a simple weight to the Frobenius\nnorm, we propose a weighted low rank (WLR) method to avoid the often\ncomputationally expensive algorithms relying on the $\\ell_1$ norm. As a proof\nof concept, a background estimation model has been presented and compared with\ntwo $\\ell_1$ norm minimization algorithms. We illustrate that as long as a\nsimple weight matrix is inferred from the data, one can use the weighted\nFrobenius norm and achieve the same or better performance.\n",
"title": "Weighted Low Rank Approximation for Background Estimation Problems"
} | null | null | null | null | true | null | 885 | null | Default | null | null |
null | {
"abstract": " In antiferromagnets, the Dzyaloshinskii-Moriya interaction lifts the\ndegeneracy of left- and right-circularly polarized spin waves. This\nrelativistic coupling increases the efficiency of spin-wave-induced domain wall\nmotion and leads to higher drift velocities. We show that in biaxial\nantiferromagnets, the spin-wave helicity controls both the direction and\nmagnitude of the magnonic force on chiral domain walls. By contrast, in\nuniaxial antiferromagnets, the magnonic force is propulsive with a helicity\ndependent strength.\n",
"title": "Controlling Chiral Domain Walls in Antiferromagnets Using Spin-Wave Helicity"
} | null | null | null | null | true | null | 886 | null | Default | null | null |
null | {
"abstract": " In disordered elastic systems, driven by displacing a parabolic confining\npotential adiabatically slowly, all advance of the system is in bursts, termed\navalanches. Avalanches have a finite extension in time, which is much smaller\nthan the waiting-time between them. Avalanches also have a finite extension\n$\\ell$ in space, i.e. only a part of the interface of size $\\ell$ moves during\nan avalanche. Here we study their spatial shape $\\left< S(x)\\right>_{\\ell}$\ngiven $\\ell$, as well as its fluctuations encoded in the second cumulant\n$\\left< S^{2}(x)\\right>_{\\ell}^{\\rm c}$. We establish scaling relations\ngoverning the behavior close to the boundary. We then give analytic results for\nthe Brownian force model, in which the microscopic disorder for each degree of\nfreedom is a random walk. Finally, we confirm these results with numerical\nsimulations. To do this properly we elucidate the influence of discretization\neffects, which also confirms the assumptions entering into the scaling ansatz.\nThis allows us to reach the scaling limit already for avalanches of moderate\nsize. We find excellent agreement for the universal shape, its fluctuations,\nincluding all amplitudes.\n",
"title": "The Spatial Shape of Avalanches"
} | null | null | null | null | true | null | 887 | null | Default | null | null |
null | {
"abstract": " The search for a superconductor with non-s-wave pairing is important not only\nfor understanding unconventional mechanisms of superconductivity but also for\nfinding new types of quasiparticles such as Majorana bound states. Materials\nwith both topological band structure and superconductivity are promising\ncandidates as $p+ip$ superconducting states can be generated through pairing\nthe spin-polarized topological surface states. In this work, the electronic and\nphonon properties of the superconductor molybdenum carbide (MoC) are studied\nwith first-principles methods. Our calculations show that nontrivial band\ntopology and superconductivity coexist in both structural phases of MoC,\nnamely, the cubic $\\alpha$ and hexagonal $\\gamma$ phases. The $\\alpha$ phase is\na strong topological insulator and the $\\gamma$ phase is a topological nodal\nline semimetal with drumhead surface states. In addition, hole doping can\nstabilize the crystal structure of the $\\alpha$ phase and elevate the\ntransition temperature in the $\\gamma$ phase. Therefore, MoC in different\nstructural forms can be a practical material platform for studying topological\nsuperconductivity and elusive Majorana fermions.\n",
"title": "Multiple Topological Electronic Phases in Superconductor MoC"
} | null | null | null | null | true | null | 888 | null | Default | null | null |
null | {
"abstract": " The extension complexity $\\mathsf{xc}(P)$ of a polytope $P$ is the minimum\nnumber of facets of a polytope that affinely projects to $P$. Let $G$ be a\nbipartite graph with $n$ vertices, $m$ edges, and no isolated vertices. Let\n$\\mathsf{STAB}(G)$ be the convex hull of the stable sets of $G$. It is easy to\nsee that $n \\leqslant \\mathsf{xc} (\\mathsf{STAB}(G)) \\leqslant n+m$. We improve\nboth of these bounds. For the upper bound, we show that $\\mathsf{xc}\n(\\mathsf{STAB}(G))$ is $O(\\frac{n^2}{\\log n})$, which is an improvement when\n$G$ has quadratically many edges. For the lower bound, we prove that\n$\\mathsf{xc} (\\mathsf{STAB}(G))$ is $\\Omega(n \\log n)$ when $G$ is the\nincidence graph of a finite projective plane. We also provide examples of\n$3$-regular bipartite graphs $G$ such that the edge vs stable set matrix of $G$\nhas a fooling set of size $|E(G)|$.\n",
"title": "Extension complexity of stable set polytopes of bipartite graphs"
} | null | null | null | null | true | null | 889 | null | Default | null | null |
null | {
"abstract": " This work provides a comprehensive scaling law based performance analysis for\nmulti-cell multi-user massive multiple-input-multiple-output (MIMO) downlink\nsystems. Imperfect channel state information (CSI), pilot contamination, and\nchannel spatial correlation are all considered. First, a sum- rate lower bound\nis derived by exploiting the asymptotically deterministic property of the\nreceived signal power, while keeping the random nature of other components in\nthe signal-to-interference-plus-noise-ratio (SINR) intact. Via a general\nscaling model on important network parameters, including the number of users,\nthe channel training energy and the data transmission power, with respect to\nthe number of base station antennas, the asymptotic scaling law of the\neffective SINR is obtained, which reveals quantitatively the tradeoff of the\nnetwork parameters. More importantly, pilot contamination and pilot\ncontamination elimination (PCE) are considered in the analytical framework. In\naddition, the applicability of the derived asymptotic scaling law in practical\nsystems with large but finite antenna numbers are discussed. Finally,\nsufficient conditions on the parameter scalings for the SINR to be\nasymptotically deterministic in the sense of mean square convergence are\nprovided, which covers existing results on such analysis as special cases and\nshows the effect of PCE explicitly.\n",
"title": "Performance Scaling Law for Multi-Cell Multi-User Massive MIMO"
} | null | null | null | null | true | null | 890 | null | Default | null | null |
null | {
"abstract": " In view of a resurgence of concern about the measurement problem, it is\npointed out that the Relativistic Transactional Interpretation (RTI) remedies\nissues previously considered as drawbacks or refutations of the original TI.\nSpecifically, once one takes into account relativistic processes that are not\nrepresentable at the non-relativistic level (such as particle creation and\nannihilation, and virtual propagation), absorption is quantitatively defined in\nunambiguous physical terms. In addition, specifics of the relativistic\ntransactional model demonstrate that the Maudlin `contingent absorber'\nchallenge to the original TI cannot even be mounted: basic features of\nestablished relativistic field theories (in particular, the asymmetry between\nfield sources and the bosonic fields, and the fact that slow-moving bound\nstates, such as atoms, are not offer waves) dictate that the `slow-moving offer\nwave' required for the challenge scenario cannot exist. It is concluded that\nissues previously considered obstacles for TI are no longer legitimately viewed\nas such, and that reconsideration of the transactional picture is warranted in\nconnection with solving the measurement problem.\n",
"title": "On the Status of the Measurement Problem: Recalling the Relativistic Transactional Interpretation"
} | null | null | [
"Physics"
]
| null | true | null | 891 | null | Validated | null | null |
null | {
"abstract": " We explore the response of Ir $5d$ orbitals to pressure in\n$\\beta$-$\\mathrm{Li_2IrO_3}$, a hyperhoneycomb iridate in proximity to a Kitaev\nquantum spin liquid (QSL) ground state. X-ray absorption spectroscopy reveals a\nreconstruction of the electronic ground state below 2 GPa, the same pressure\nrange where x-ray magnetic circular dichroism shows an apparent collapse of\nmagnetic order. The electronic reconstruction, which manifests a reduction in\nthe effective spin-orbit (SO) interaction in $5d$ orbitals, pushes\n$\\beta$-$\\mathrm{Li_2IrO_3}$ further away from the pure $J_{\\rm eff}=1/2$\nlimit. Although lattice symmetry is preserved across the electronic transition,\nx-ray diffraction shows a highly anisotropic compression of the hyperhoneycomb\nlattice which affects the balance of bond-directional Ir-Ir exchange\ninteractions driven by spin-orbit coupling at Ir sites. An enhancement of\nsymmetric anisotropic exchange over Kitaev and Heisenberg exchange interactions\nseen in theoretical calculations that use precisely this anisotropic Ir-Ir bond\ncompression provides one possible route to realization of a QSL state in this\nhyperhoneycomb iridate at high pressures.\n",
"title": "Pressure-tuning of bond-directional exchange interactions and magnetic frustration in hyperhoneycomb iridate $β$-$\\mathrm{Li_2IrO_3}$"
} | null | null | null | null | true | null | 892 | null | Default | null | null |
null | {
"abstract": " An infinite convergent sum of independent and identically distributed random\nvariables discounted by a multiplicative random walk is called perpetuity,\nbecause of a possible actuarial application. We give three disjoint groups of\nsufficient conditions which ensure that the distribution right tail of a\nperpetuity $\\mathbb{P}\\{X>x\\}$ is asymptotic to $ax^ce^{-bx}$ as $x\\to\\infty$\nfor some $a,b>0$ and $c\\in\\mathbb{R}$. Our results complement those of Denisov\nand Zwart [J. Appl. Probab. 44 (2007), 1031--1046]. As an auxiliary tool we\nprovide criteria for the finiteness of the one-sided exponential moments of\nperpetuities. Several examples are given in which the distributions of\nperpetuities are explicitly identified.\n",
"title": "On perpetuities with gamma-like tails"
} | null | null | null | null | true | null | 893 | null | Default | null | null |
null | {
"abstract": " We consider the problem of estimating from sample paths the absolute spectral\ngap $\\gamma_*$ of a reversible, irreducible and aperiodic Markov chain\n$(X_t)_{t \\in \\mathbb{N}}$ over a finite state $\\Omega$. We propose the ${\\tt\nUCPI}$ (Upper Confidence Power Iteration) algorithm for this problem, a\nlow-complexity algorithm which estimates the spectral gap in time ${\\cal O}(n)$\nand memory space ${\\cal O}((\\ln n)^2)$ given $n$ samples. This is in stark\ncontrast with most known methods which require at least memory space ${\\cal\nO}(|\\Omega|)$, so that they cannot be applied to large state spaces.\nFurthermore, ${\\tt UCPI}$ is amenable to parallel implementation.\n",
"title": "Computationally Efficient Estimation of the Spectral Gap of a Markov Chain"
} | null | null | null | null | true | null | 894 | null | Default | null | null |
null | {
"abstract": " The state-of-the-art (SOTA) for mixed precision training is dominated by\nvariants of low precision floating point operations, and in particular, FP16\naccumulating into FP32 Micikevicius et al. (2017). On the other hand, while a\nlot of research has also happened in the domain of low and mixed-precision\nInteger training, these works either present results for non-SOTA networks (for\ninstance only AlexNet for ImageNet-1K), or relatively small datasets (like\nCIFAR-10). In this work, we train state-of-the-art visual understanding neural\nnetworks on the ImageNet-1K dataset, with Integer operations on General Purpose\n(GP) hardware. In particular, we focus on Integer Fused-Multiply-and-Accumulate\n(FMA) operations which take two pairs of INT16 operands and accumulate results\ninto an INT32 output.We propose a shared exponent representation of tensors and\ndevelop a Dynamic Fixed Point (DFP) scheme suitable for common neural network\noperations. The nuances of developing an efficient integer convolution kernel\nis examined, including methods to handle overflow of the INT32 accumulator. We\nimplement CNN training for ResNet-50, GoogLeNet-v1, VGG-16 and AlexNet; and\nthese networks achieve or exceed SOTA accuracy within the same number of\niterations as their FP32 counterparts without any change in hyper-parameters\nand with a 1.8X improvement in end-to-end training throughput. To the best of\nour knowledge these results represent the first INT16 training results on GP\nhardware for ImageNet-1K dataset using SOTA CNNs and achieve highest reported\naccuracy using half-precision\n",
"title": "Mixed Precision Training of Convolutional Neural Networks using Integer Operations"
} | null | null | null | null | true | null | 895 | null | Default | null | null |
null | {
"abstract": " The foreseen implementations of the Small Size Telescopes (SST) in CTA will\nprovide unique insights into the highest energy gamma rays offering fundamental\nmeans to discover and under- stand the sources populating the Galaxy and our\nlocal neighborhood. Aiming at such a goal, the SST-1M is one of the three\ndifferent implementations that are being prototyped and tested for CTA. SST-1M\nis a Davies-Cotton single mirror telescope equipped with a unique camera\ntechnology based on SiPMs with demonstrated advantages over classical\nphotomultipliers in terms of duty-cycle. In this contribution, we describe the\ntelescope components, the camera, and the trigger and readout system. The\nresults of the commissioning of the camera using a dedicated test setup are\nthen presented. The performances of the camera first prototype in terms of\nexpected trigger rates and trigger efficiencies for different night-sky\nbackground conditions are presented, and the camera response is compared to\nend-to-end simulations.\n",
"title": "Performance of a small size telescope (SST-1M) camera for gamma-ray astronomy with the Cherenkov Telescope Array"
} | null | null | [
"Physics"
]
| null | true | null | 896 | null | Validated | null | null |
null | {
"abstract": " We obtain bounded for all $t$ solutions of ordinary differential equations as\nlimits of the solutions of the corresponding Dirichlet problems on $(-L,L)$,\nwith $L \\rightarrow \\infty$. We derive a priori estimates for the Dirichlet\nproblems, allowing passage to the limit, via a diagonal sequence. This approach\ncarries over to the PDE case.\n",
"title": "Bounded solutions for a class of Hamiltonian systems"
} | null | null | null | null | true | null | 897 | null | Default | null | null |
null | {
"abstract": " Waveforms of gravitational waves provide information about a variety of\nparameters for the binary system merging. However, standard calculations have\nbeen performed assuming a FLRW universe with no perturbations. In reality this\nassumption should be dropped: we show that the inclusion of cosmological\nperturbations translates into corrections to the estimate of astrophysical\nparameters derived for the merging binary systems. We compute corrections to\nthe estimate of the luminosity distance due to velocity, volume, lensing and\ngravitational potential effects. Our results show that the amplitude of the\ncorrections will be negligible for current instruments, mildly important for\nexperiments like the planned DECIGO, and very important for future ones such as\nthe Big Bang Observer.\n",
"title": "Cosmological perturbation effects on gravitational-wave luminosity distance estimates"
} | null | null | [
"Physics"
]
| null | true | null | 898 | null | Validated | null | null |
null | {
"abstract": " A simple DNA-based data storage scheme is demonstrated in which information\nis written using \"addressing\" oligonucleotides. In contrast to other methods\nthat allow arbitrary code to be stored, the resulting DNA is suitable for\ndownstream enzymatic and biological processing. This capability is crucial for\nDNA computers, and may allow for a diverse array of computational operations to\nbe carried out using this DNA. Although here we use gel-based methods for\ninformation readout, we also propose more advanced methods involving\nprotein/DNA complexes and atomic force microscopy/nano-pore schemes for data\nreadout.\n",
"title": "Biocompatible Writing of Data into DNA"
} | null | null | null | null | true | null | 899 | null | Default | null | null |
null | {
"abstract": " This paper presents VEC-NBT, a variation on the unsupervised graph clustering\ntechnique VEC, which improves upon the performance of the original algorithm\nsignificantly for sparse graphs. VEC employs a novel application of the\nstate-of-the-art word2vec model to embed a graph in Euclidean space via random\nwalks on the nodes of the graph. In VEC-NBT, we modify the original algorithm\nto use a non-backtracking random walk instead of the normal backtracking random\nwalk used in VEC. We introduce a modification to a non-backtracking random\nwalk, which we call a begrudgingly-backtracking random walk, and show\nempirically that using this model of random walks for VEC-NBT requires shorter\nwalks on the graph to obtain results with comparable or greater accuracy than\nVEC, especially for sparser graphs.\n",
"title": "Faster Clustering via Non-Backtracking Random Walks"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 900 | null | Validated | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.