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": " It is shown that for controlled Moran constructions in $\\mathbb{R}$,\nincluding the (sub) self-similar and more generally, (sub) self-conformal sets,\nthe quasi-Assouad dimension coincides with the upper box dimension. This can be\nextended to some special classes of self-similar sets in higher dimensions. The\nconnections between quasi-Assouad dimension and tangents are studied. We show\nthat sets with decreasing gaps have quasi-Assouad dimension $0$ or $1$ and we\nexhibit an example of a set in the plane whose quasi-Assouad dimension is\nsmaller than that of its projection onto the $x$-axis, showing that\nquasi-Assouad dimension may increase under Lipschitz mappings.\n",
"title": "Properties of Quasi-Assouad dimension"
}
| null | null | null | null | true | null |
14201
| null |
Default
| null | null |
null |
{
"abstract": " Predictable Feature Analysis (PFA) (Richthofer, Wiskott, ICMLA 2015) is an\nalgorithm that performs dimensionality reduction on high dimensional input\nsignal. It extracts those subsignals that are most predictable according to a\ncertain prediction model. We refer to these extracted signals as predictable\nfeatures.\nIn this work we extend the notion of PFA to take supplementary information\ninto account for improving its predictions. Such information can be a\nmultidimensional signal like the main input to PFA, but is regarded external.\nThat means it won't participate in the feature extraction - no features get\nextracted or composed of it. Features will be exclusively extracted from the\nmain input such that they are most predictable based on themselves and the\nsupplementary information. We refer to this enhanced PFA as PFAx (PFA\nextended).\nEven more important than improving prediction quality is to observe the\neffect of supplementary information on feature selection. PFAx transparently\nprovides insight how the supplementary information adds to prediction quality\nand whether it is valuable at all. Finally we show how to invert that relation\nand can generate the supplementary information such that it would yield a\ncertain desired outcome of the main signal.\nWe apply this to a setting inspired by reinforcement learning and let the\nalgorithm learn how to control an agent in an environment. With this method it\nis feasible to locally optimize the agent's state, i.e. reach a certain goal\nthat is near enough. We are preparing a follow-up paper that extends this\nmethod such that also global optimization is feasible.\n",
"title": "PFAx: Predictable Feature Analysis to Perform Control"
}
| null | null | null | null | true | null |
14202
| null |
Default
| null | null |
null |
{
"abstract": " Elements composing complex systems usually interact in several different ways\nand as such the interaction architecture is well modelled by a multiplex\nnetwork. However often this architecture is hidden, as one usually only has\nexperimental access to an aggregated projection. A fundamental challenge is\nthus to determine whether the hidden underlying architecture of complex systems\nis better modelled as a single interaction layer or results from the\naggregation and interplay of multiple layers. Here we show that using local\ninformation provided by a random walker navigating the aggregated network one\ncan decide in a robust way if the underlying structure is a multiplex or not\nand, in the former case, to determine the most probable number of hidden\nlayers. As a byproduct, we show that the mathematical formalism also provides a\nprincipled solution for the optimal decomposition and projection of complex,\nnon-Markovian dynamics into a Markov switching combination of diffusive modes.\nWe validate the proposed methodology with numerical simulations of both (i)\nrandom walks navigating hidden multiplex networks (thereby reconstructing the\ntrue hidden architecture) and (ii) Markovian and non-Markovian continuous\nstochastic processes (thereby reconstructing an effective multiplex\ndecomposition where each layer accounts for a different diffusive mode). We\nalso state and prove two existence theorems guaranteeing that an exact\nreconstruction of the dynamics in terms of these hidden jump-Markov models is\nalways possible for arbitrary finite-order Markovian and fully non-Markovian\nprocesses. Finally, we showcase the applicability of the method to experimental\nrecordings from (i) the mobility dynamics of human players in an online\nmultiplayer game and (ii) the dynamics of RNA polymerases at the\nsingle-molecule level.\n",
"title": "Multiplex decomposition of non-Markovian dynamics and the hidden layer reconstruction problem"
}
| null | null | null | null | true | null |
14203
| null |
Default
| null | null |
null |
{
"abstract": " Many studies show that the acquisition of knowledge is the key to build\ncompetitive advantage of companies. We propose a simple model of knowledge\ntransfer within the organization and we implement the proposed model using\ncellular automata technique. In this paper the organisation is considered in\nthe context of complex systems. In this perspective, the main role in\norganisation is played by the network of informal contacts and the distributed\nleadership. The goal of this paper is to check which factors influence the\nefficiency and effectiveness of knowledge transfer. Our studies indicate a\nsignificant role of initial concentration of chunks of knowledge for knowledge\ntransfer process, and the results suggest taking action in the organisation to\nshorten the distance (social distance) between people with different levels of\nknowledge, or working out incentives to share knowledge.\n",
"title": "Model of knowledge transfer within an organisation"
}
| null | null | null | null | true | null |
14204
| null |
Default
| null | null |
null |
{
"abstract": " Uncertainty propagation of large scale discrete supply chains can be\nprohibitive when a large number of events occur during the simulated period and\ndiscrete event simulations (DES) are costly. We present a time bucket method to\napproximate and accelerate the DES of supply chains. Its stochastic version,\nwhich we call the L(logistic)-leap method, can be viewed as an extension of the\nleap methods, e.g., tau-leap, D-leap, developed in the chemical engineering\ncommunity for the acceleration of stochastic DES of chemical reactions. The\nL-leap method instantaneously updates the system state vector at discrete time\npoints and the production rates and policies of a supply chain are assumed to\nbe stationary during each time bucket. We propose to use Multilevel Monte Carlo\n(MLMC) to efficiently propagate the uncertainties in a supply chain network,\nwhere the levels are naturally defined by the sizes of the time buckets of the\nsimulations. We demonstrate the efficiency and accuracy of our methods using\nfour numerical examples derived from a real world manufacturing material flow.\nIn these examples, our multilevel L-leap approach can be faster than the\nstandard Monte Carlo (MC) method by one or two orders of magnitudes without\ncompromising the accuracy.\n",
"title": "Efficient Propagation of Uncertainties in Manufacturing Supply Chains: Time Buckets, L-leap and Multilevel Monte Carlo"
}
| null | null |
[
"Statistics"
] | null | true | null |
14205
| null |
Validated
| null | null |
null |
{
"abstract": " Location fingerprinting locates devices based on pattern matching signal\nobservations to a pre-defined signal map. This paper introduces a technique to\nenable fast signal map creation given a dedicated surveyor with a smartphone\nand floorplan. Our technique (PFSurvey) uses accelerometer, gyroscope and\nmagnetometer data to estimate the surveyor's trajectory post-hoc using\nSimultaneous Localisation and Mapping and particle filtering to incorporate a\nbuilding floorplan. We demonstrate conventional methods can fail to recover the\nsurvey path robustly and determine the room unambiguously. To counter this we\nuse a novel loop closure detection method based on magnetic field signals and\npropose to incorporate the magnetic loop closures and straight-line constraints\ninto the filtering process to ensure robust trajectory recovery. We show this\nallows room ambiguities to be resolved.\nAn entire building can be surveyed by the proposed system in minutes rather\nthan days. We evaluate in a large office space and compare to state-of-the-art\napproaches. We achieve trajectories within 1.1 m of the ground truth 90% of the\ntime. Output signal maps well approximate those built from conventional,\nlaborious manual survey. We also demonstrate that the signal maps built by\nPFSurvey provide similar or even better positioning performance than the manual\nsignal maps.\n",
"title": "Semi-automated Signal Surveying Using Smartphones and Floorplans"
}
| null | null | null | null | true | null |
14206
| null |
Default
| null | null |
null |
{
"abstract": " We prove an explicit formula for the first non-zero entry in the n-th row of\nthe graded Betti table of an n-dimensional projective toric variety associated\nto a normal polytope with at least one interior lattice point. This applies to\nVeronese embeddings of projective space where we prove a special case of a\nconjecture of Ein and Lazarsfeld. We also prove an explicit formula for the\nentire n-th row when the interior of the polytope is one-dimensional. All\nresults are valid over an arbitrary field k.\n",
"title": "On the n-th row of the graded Betti table of an n-dimensional toric variety"
}
| null | null | null | null | true | null |
14207
| null |
Default
| null | null |
null |
{
"abstract": " We will characterize topologically conjugate two-sided topological Markov\nshifts $(\\bar{X}_A,\\bar{\\sigma}_A)$ in terms of the associated asymptotic\nRuelle $C^*$-algebras ${\\mathcal{R}}_A$ with its commutative $C^*$-subalgebras\n$C(\\bar{X}_A)$ and the canonical circle actions. We will also show that\nextended Ruelle algebras ${\\widetilde{\\mathcal{R}}}_A$, which are purely\ninfinite version of the asymptotic Ruelle algebras, with its commutative\n$C^*$-subalgebras $C(\\bar{X}_A)$ and the canonical torus actions $\\gamma^A$ are\ncomplete invariants for topological conjugacy of two-sided topological Markov\nshifts. We then have a computable topological conjugacy invariant, written in\nterms of the underlying matrix, of a two-sided topological Markov shift by\nusing K-theory of the extended Ruelle algebra. The diagonal action of\n$\\gamma^A$ has a unique KMS-state on ${\\widetilde{\\mathcal{R}}}_A$, which is an\nextension of the Parry measure on $\\bar{X}_A$.\n",
"title": "Topological conjugacy of topological Markov shifts and Ruelle algebras"
}
| null | null | null | null | true | null |
14208
| null |
Default
| null | null |
null |
{
"abstract": " Encrypted database systems provide a great method for protecting sensitive\ndata in untrusted infrastructures. These systems are built using either\nspecial-purpose cryptographic algorithms that support operations over encrypted\ndata, or by leveraging trusted computing co-processors. Strong cryptographic\nalgorithms usually result in high performance overheads (e.g., public-key\nencryptions, garbled circuits), while weaker algorithms (e.g., order-preserving\nencryption) result in large leakage profiles. On the other hand, some encrypted\ndatabase systems (e.g., Cipherbase, TrustedDB) leverage non-standard trusted\ncomputing devices, and are designed to work around their specific architectural\nlimitations.\nIn this work we build StealthDB -- an encrypted database system from Intel\nSGX. Our system can run on any newer generation Intel CPU. StealthDB has a very\nsmall trusted computing base, scales to large datasets, requires no DBMS\nchanges, and provides strong security guarantees at steady state and during\nquery execution.\n",
"title": "StealthDB: a Scalable Encrypted Database with Full SQL Query Support"
}
| null | null | null | null | true | null |
14209
| null |
Default
| null | null |
null |
{
"abstract": " The main aim of this survey paper is to gather together some results\nconcerning the Calabi type duality discovered by Hojoo Lee between certain\nfamilies of (spacelike) graphs with constant mean curvature in Riemannian and\nLorentzian homogeneous 3-manifolds with isometry group of dimension 4. The\nduality is conformal and swaps mean curvature and bundle curvature, and we will\nrevisit it by giving a more general statement in terms of conformal immersions.\nThis will show that some features in the theory of surfaces with mean curvature\n$\\frac{1}{2}$ in $\\mathbb{H}^2\\times\\mathbb{R}$ or minimal surfaces in the\nHeisenberg space have nice geometric interpretations in terms of their dual\nLorentzian counterparts. We will briefly discuss some applications such as\ngradient estimates for entire minimal graphs in Heisenberg space or the\nexistence of complete spacelike surfaces, and we will also give an uniform\ntreatment to the behavior of the duality with respect to ambient isometries.\nFinally, some open questions are posed in the last section.\n",
"title": "On the conformal duality between constant mean curvature surfaces in $\\mathbb{E}(κ,τ)$ and $\\mathbb{L}(κ,τ)$"
}
| null | null | null | null | true | null |
14210
| null |
Default
| null | null |
null |
{
"abstract": " Contact-rich manipulation tasks in unstructured environments often require\nboth haptic and visual feedback. However, it is non-trivial to manually design\na robot controller that combines modalities with very different\ncharacteristics. While deep reinforcement learning has shown success in\nlearning control policies for high-dimensional inputs, these algorithms are\ngenerally intractable to deploy on real robots due to sample complexity. We use\nself-supervision to learn a compact and multimodal representation of our\nsensory inputs, which can then be used to improve the sample efficiency of our\npolicy learning. We evaluate our method on a peg insertion task, generalizing\nover different geometry, configurations, and clearances, while being robust to\nexternal perturbations. Results for simulated and real robot experiments are\npresented.\n",
"title": "Making Sense of Vision and Touch: Self-Supervised Learning of Multimodal Representations for Contact-Rich Tasks"
}
| null | null | null | null | true | null |
14211
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\Omega:=\\left( a,b\\right) \\subset\\mathbb{R}$, $m\\in L^{1}\\left(\n\\Omega\\right) $ and $\\lambda>0$ be a real parameter. Let $\\mathcal{L}$ be the\ndifferential operator given by $\\mathcal{L}u:=-\\phi\\left( u^{\\prime}\\right)\n^{\\prime}+r\\left( x\\right) \\phi\\left( u\\right) $, where $\\phi\n:\\mathbb{R\\rightarrow R}$ is an odd increasing homeomorphism and $0\\leq r\\in\nL^{1}\\left( \\Omega\\right) $. We study the existence of positive solutions for\nproblems of the form $\\mathcal{L}u=\\lambda m\\left( x\\right) f\\left( u\\right)$\nin $\\Omega,$ $u=0$ on $\\partial\\Omega$, where $f:\\left[ 0,\\infty\\right)\n\\rightarrow\\left[ 0,\\infty\\right) $ is a continuos function which is, roughly\nspeaking, sublinear with respect to $\\phi$. Our approach combines the sub and\nsupersolution method with some estimates on related nonlinear problems. We\npoint out that our results are new even in the cases $r\\equiv0$ and/or\n$m\\geq0$.\n",
"title": "Positive solutions for nonlinear problems involving the one-dimensional ϕ-Laplacian"
}
| null | null | null | null | true | null |
14212
| null |
Default
| null | null |
null |
{
"abstract": " We study the structure and stability of vortex lattices in two-component\nrotating Bose-Einstein condensates with intrinsic dipole-dipole interactions\n(DDIs) and contact interactions. To address experimentally accessible coupled\nsystems, we consider $^{164}$Dy-$^{162}$Dy and $^{168}$Er-$^{164}$Dy mixtures,\nwhich feature different miscibilities. The corresponding dipole moments are\n$\\mu_{\\mathrm{Dy}}=10\\mu_{\\mathrm{B}}$ and $\\mu_{\\mathrm{Er}}=\n7\\mu_{\\mathrm{B}}$, where $\\mu_{\\mathrm{B}}$ is the Bohr magneton. For\ncomparison, we also discuss a case where one of the species is non dipolar.\nUnder a large aspect ratio of the trap, we consider mixtures in the\npancake-shaped format, which are modeled by effective two-dimensional coupled\nGross-Pitaevskii equations, with a fixed polarization of the magnetic dipoles.\nThen, the miscibility and vortex-lattice structures are studied, by varying the\ncoefficients of the contact interactions (assuming the use of the\nFeshbach-resonance mechanism) and the rotation frequency. We present phase\ndiagrams for several types of lattices in the parameter plane of the rotation\nfrequency and ratio of inter- and intra-species scattering lengths. The vortex\nstructures are found to be diverse for the more miscible $^{164}$Dy-$^{162}$Dy\nmixture, with a variety of shapes, whereas, for the less miscible case of\n$^{168}$Er-$^{164}$Dy, the lattice patterns mainly feature circular or square\nformats.\n",
"title": "Vortex lattices in binary Bose-Einstein condensates with dipole-dipole interactions"
}
| null | null |
[
"Physics"
] | null | true | null |
14213
| null |
Validated
| null | null |
null |
{
"abstract": " Estimating the tail index parameter is one of the primal objectives in\nextreme value theory. For heavy-tailed distributions the Hill estimator is the\nmost popular way to estimate the tail index parameter. Improving the Hill\nestimator was aimed by recent works with different methods, for example by\nusing bootstrap, or Kolmogorov-Smirnov metric. These methods are asymptotically\nconsistent, but for tail index $\\xi >1$ and smaller sample sizes the estimation\nfails to approach the theoretical value for realistic sample sizes. In this\npaper, we introduce a new empirical method, which can estimate high tail index\nparameters well and might also be useful for relatively small sample sizes.\n",
"title": "Regression estimator for the tail index"
}
| null | null | null | null | true | null |
14214
| null |
Default
| null | null |
null |
{
"abstract": " Metamaterial analogues of electromagnetically induced transparency (EIT) have\nbeen intensively studied and widely employed for slow light and enhanced\nnonlinear effects. In particular, the active modulation of the EIT analogue and\nwell-controlled group delay in metamaterials have shown great prospects in\noptical communication networks. Previous studies have focused on the optical\ncontrol of the EIT analogue by integrating the photoactive materials into the\nunit cell, however, the response time is limited by the recovery time of the\nexcited carriers in these bulk materials. Graphene has recently emerged as an\nexceptional optoelectronic material. It shows an ultrafast relaxation time on\nthe order of picosecond and its conductivity can be tuned via manipulating the\nFermi energy. Here we integrate a monolayer graphene into metal-based terahertz\n(THz) metamaterials, and realize a complete modulation in the resonance\nstrength of the EIT analogue at the accessible Fermi energy. The physical\nmechanism lies in the active tuning the damping rate of the dark mode resonator\nthrough the recombination effect of the conductive graphene. Note that the\nmonolayer morphology in our work is easier to fabricate and manipulate than\nisolated fashion. This work presents a novel modulation strategy of the EIT\nanalogue in the hybrid metamaterials, and pave the way towards designing very\ncompact slow light devices to meet future demand of ultrafast optical signal\nprocessing.\n",
"title": "Active modulation of electromagnetically induced transparency analogue in terahertz hybrid metal-graphene metamaterials"
}
| null | null | null | null | true | null |
14215
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study a special case of the completion of cusp\nKähler-Einstein metric on the regular part of varieties by taking the\ncontinuity method proposed by La Nave and Tian. The differential geometric and\nalgebro-geometric properties of the noncollapsing limit in the continuity\nmethod with cusp singularities will be investigated.\n",
"title": "The continuity equation with cusp singularities"
}
| null | null | null | null | true | null |
14216
| null |
Default
| null | null |
null |
{
"abstract": " For $d\\geq1$, we study the simplicial structure of the chain complex\nassociated to the higher order Hochschild homology over the $d$-sphere. We\ndiscuss $H_\\bullet^{S^d}(A,M)$ by way of a bar-like resolution\n$\\mathcal{B}^d(A)$ in the context of simplicial modules. Besides the general\ncase, we give explicit detail corresponding to $S^3$. We also present a\ndescription of what can replace these bar-like resolutions in order to aid with\ncomputation. The cohomology version can be done following a similar\nconstruction, of which we make mention.\n",
"title": "Simplicial Structures for Higher Order Hochschild Homology over the $d$-Sphere"
}
| null | null | null | null | true | null |
14217
| null |
Default
| null | null |
null |
{
"abstract": " In some problems there is information about the destination of a moving\nobject. An example is an airliner flying from an origin to a destination. Such\nproblems have three main components: an origin, a destination, and motion in\nbetween. To emphasize that the motion trajectories end up at the destination,\nwe call them \\textit{destination-directed trajectories}. The Markov sequence is\nnot flexible enough to model such trajectories. Given an initial density and an\nevolution law, the future of a Markov sequence is determined probabilistically.\nOne class of conditionally Markov (CM) sequences, called the $CM_L$ sequence\n(including the Markov sequence as a special case), has the following main\ncomponents: a joint endpoint density (i.e., an initial density and a final\ndensity conditioned on the initial) and a Markov-like evolution law. This paper\nproposes using the $CM_L$ sequence for modeling destination-directed\ntrajectories. It is demonstrated how the $CM_L$ sequence enjoys several\ndesirable properties for destination-directed trajectory modeling. Some\nsimulations of trajectory modeling and prediction are presented for\nillustration.\n",
"title": "Destination-Directed Trajectory Modeling and Prediction Using Conditionally Markov Sequences"
}
| null | null | null | null | true | null |
14218
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the extension of Monadic Second Order logic, interpreted over\ninfinite words and trees, with generalized \"for almost all\" quantifiers\ninterpreted using the notions of Baire category and Lebesgue measure.\n",
"title": "Monadic Second Order Logic with Measure and Category Quantifiers"
}
| null | null | null | null | true | null |
14219
| null |
Default
| null | null |
null |
{
"abstract": " For the constant-stress layer of wall turbulence, two-point correlations of\nvelocity fluctuations are studied theoretically by using the attached-eddy\nhypothesis, i.e., a phenomenological model of a random superposition of\nenergy-containing eddies that are attached to the wall. While the previous\nstudies had invoked additional assumptions, we focus on the minimum assumptions\nof the hypothesis to derive its most general forms of the correlation\nfunctions. They would allow us to use or assess the hypothesis without any\neffect of those additional assumptions. We also study the energy spectra and\nthe two-point correlations of the rate of momentum transfer and of the rate of\nenergy dissipation.\n",
"title": "Two-point correlation in wall turbulence according to the attached-eddy hypothesis"
}
| null | null | null | null | true | null |
14220
| null |
Default
| null | null |
null |
{
"abstract": " We consider a class of variational problems for densities that repel each\nother at distance. Typical examples are given by the Dirichlet functional and\nthe Rayleigh functional \\[\nD(\\mathbf{u}) = \\sum_{i=1}^k \\int_{\\Omega} |\\nabla u_i|^2 \\quad \\text{or}\n\\quad R(\\mathbf{u}) = \\sum_{i=1}^k \\frac{\\int_{\\Omega} |\\nabla\nu_i|^2}{\\int_{\\Omega} u_i^2} \\] minimized in the class of\n$H^1(\\Omega,\\mathbb{R}^k)$ functions attaining some boundary conditions on\n$\\partial \\Omega$, and subjected to the constraint \\[\n\\mathrm{dist} (\\{u_i > 0\\}, \\{u_j > 0\\}) \\ge 1 \\qquad \\forall i \\neq j. \\]\nFor these problems, we investigate the optimal regularity of the solutions,\nprove a free-boundary condition, and derive some preliminary results\ncharacterizing the free boundary $\\partial \\{\\sum_{i=1}^k u_i > 0\\}$.\n",
"title": "Variational problems with long-range interaction"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14221
| null |
Validated
| null | null |
null |
{
"abstract": " This paper presents a comparison of six machine learning (ML) algorithms:\nGRU-SVM (Agarap, 2017), Linear Regression, Multilayer Perceptron (MLP), Nearest\nNeighbor (NN) search, Softmax Regression, and Support Vector Machine (SVM) on\nthe Wisconsin Diagnostic Breast Cancer (WDBC) dataset (Wolberg, Street, &\nMangasarian, 1992) by measuring their classification test accuracy and their\nsensitivity and specificity values. The said dataset consists of features which\nwere computed from digitized images of FNA tests on a breast mass (Wolberg,\nStreet, & Mangasarian, 1992). For the implementation of the ML algorithms, the\ndataset was partitioned in the following fashion: 70% for training phase, and\n30% for the testing phase. The hyper-parameters used for all the classifiers\nwere manually assigned. Results show that all the presented ML algorithms\nperformed well (all exceeded 90% test accuracy) on the classification task. The\nMLP algorithm stands out among the implemented algorithms with a test accuracy\nof ~99.04%.\n",
"title": "On Breast Cancer Detection: An Application of Machine Learning Algorithms on the Wisconsin Diagnostic Dataset"
}
| null | null | null | null | true | null |
14222
| null |
Default
| null | null |
null |
{
"abstract": " Autoencoders are a deep learning model for representation learning. When\ntrained to minimize the Euclidean distance between the data and its\nreconstruction, linear autoencoders (LAEs) learn the subspace spanned by the\ntop principal directions but cannot learn the principal directions themselves.\nIn this paper, we prove that $L_2$-regularized LAEs learn the principal\ndirections as the left singular vectors of the decoder, providing an extremely\nsimple and scalable algorithm for rank-$k$ SVD. More generally, we consider\nLAEs with (i) no regularization, (ii) regularization of the composition of the\nencoder and decoder, and (iii) regularization of the encoder and decoder\nseparately. We relate the minimum of (iii) to the MAP estimate of probabilistic\nPCA and show that for all critical points the encoder and decoder are\ntransposes. Building on topological intuition, we smoothly parameterize the\ncritical manifolds for all three losses via a novel unified framework and\nillustrate these results empirically. Overall, this work clarifies the\nrelationship between autoencoders and Bayesian models and between\nregularization and orthogonality.\n",
"title": "Loss Landscapes of Regularized Linear Autoencoders"
}
| null | null | null | null | true | null |
14223
| null |
Default
| null | null |
null |
{
"abstract": " Ocean flows are routinely inferred from low-resolution satellite altimetry\nmeasurements of sea surface height assuming a geostrophic balance. Recent\nnonlinear dynamical systems techniques have revealed that surface currents\nderived from altimetry can support mesoscale eddies with material boundaries\nthat do not filament for many months, thereby representing effective transport\nmechanisms. However, the long-range Lagrangian coherence assessed for mesoscale\neddy boundaries detected from altimetry is constrained by the impossibility of\ncurrent altimeters to resolve ageostrophic submesoscale motions. These may act\nto prevent Lagrangian coherence from manifesting in the rigorous form described\nby the nonlinear dynamical systems theories. Here we use a combination of\nsatellite ocean color and surface drifter trajectory data, rarely available\nsimultaneously over an extended period of time, to provide observational\nevidence for the enduring Lagrangian coherence of a Loop Current ring detected\nfrom altimetry. We also seek indications of this behavior in the flow produced\nby a data-assimilative system which demonstrated ability to reproduce observed\nrelative dispersion statistics down into the marginally submesoscale range.\nHowever, the simulated flow, total surface and subsurface or subsampled\nemulating altimetry, is not found to support the long-lasting Lagrangian\ncoherence that characterizes the observed ring. This highlights the importance\nof the Lagrangian metrics produced by the nonlinear dynamical systems tools\nemployed here in assessing model performance.\n",
"title": "Enduring Lagrangian coherence of a Loop Current ring assessed using independent observations"
}
| null | null | null | null | true | null |
14224
| null |
Default
| null | null |
null |
{
"abstract": " A flexible approach for modeling both dynamic event counting and dynamic\nlink-based networks based on counting processes is proposed, and estimation in\nthese models is studied. We consider nonparametric likelihood based estimation\nof parameter functions via kernel smoothing. The asymptotic behavior of these\nestimators is rigorously analyzed by allowing the number of nodes to tend to\ninfinity. The finite sample performance of the estimators is illustrated\nthrough an empirical analysis of bike share data.\n",
"title": "Nonparametric inference for continuous-time event counting and link-based dynamic network models"
}
| null | null | null | null | true | null |
14225
| null |
Default
| null | null |
null |
{
"abstract": " As the core issue of blockchain, the mining requires solving a proof-of-work\npuzzle, which is resource expensive to implement in mobile devices due to high\ncomputing power needed. Thus, the development of blockchain in mobile\napplications is restricted. In this paper, we consider the edge computing as\nthe network enabler for mobile blockchain. In particular, we study optimal\npricing-based edge computing resource management to support mobile blockchain\napplications where the mining process can be offloaded to an Edge computing\nService Provider (ESP). We adopt a two-stage Stackelberg game to jointly\nmaximize the profit of the ESP and the individual utilities of different\nminers. In Stage I, the ESP sets the price of edge computing services. In Stage\nII, the miners decide on the service demand to purchase based on the observed\nprices. We apply the backward induction to analyze the sub-game perfect\nequilibrium in each stage for uniform and discriminatory pricing schemes.\nFurther, the existence and uniqueness of Stackelberg game are validated for\nboth pricing schemes. At last, the performance evaluation shows that the ESP\nintends to set the maximum possible value as the optimal price for profit\nmaximization under uniform pricing. In addition, the discriminatory pricing\nhelps the ESP encourage higher total service demand from miners and achieve\ngreater profit correspondingly.\n",
"title": "Optimal Pricing-Based Edge Computing Resource Management in Mobile Blockchain"
}
| null | null | null | null | true | null |
14226
| null |
Default
| null | null |
null |
{
"abstract": " A general conjecture is stated on the cone of automorphic vector bundles\nadmitting nonzero global sections on schemes endowed with a smooth, surjective\nmorphism to a stack of $G$-zips of connected-Hodge-type; such schemes should\ninclude all Hodge-type Shimura varieties with hyperspecial level. We prove our\nconjecture for groups of type $A_1^n$, $C_2$ and $\\mathbf F_p$-split groups of\ntype $A_2$ (this includes all Hilbert-Blumenthal varieties and should also\napply to Siegel modular threefolds and Picard modular surfaces). An example is\ngiven to show that our conjecture can fail for zip data not of\nconnected-Hodge-type.\n",
"title": "Automorphic vector bundles with global sections on $G$-${\\tt Zip}^{\\mathcal Z}$-schemes"
}
| null | null | null | null | true | null |
14227
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we propose a composition/decomposition framework for\nadversarially training generative models on composed data - data where each\nsample can be thought of as being constructed from a fixed number of\ncomponents. In our framework, samples are generated by sampling components from\ncomponent generators and feeding these components to a composition function\nwhich combines them into a \"composed sample\". This compositional training\napproach improves the modularity, extensibility and interpretability of\nGenerative Adversarial Networks (GANs) - providing a principled way to\nincrementally construct complex models out of simpler component models, and\nallowing for explicit \"division of responsibility\" between these components.\nUsing this framework, we define a family of learning tasks and evaluate their\nfeasibility on two datasets in two different data modalities (image and text).\nLastly, we derive sufficient conditions such that these compositional\ngenerative models are identifiable. Our work provides a principled approach to\nbuilding on pre-trained generative models or for exploiting the compositional\nnature of data distributions to train extensible and interpretable models.\n",
"title": "Composition and decomposition of GANs"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14228
| null |
Validated
| null | null |
null |
{
"abstract": " We establish that every $K$-quasiconformal mapping $w$ of the unit ball $\\IB$\nonto a $C^2$-Jordan domain $\\Omega$ is Hölder continuous with constant\n$\\alpha= 2-\\frac{n}{p}$, provided that its weak Laplacean $\\Delta w$ is in $\nL^p(\\IB)$ for some $n/2<p<n$. In particular it is Hölder continuous for every\n$0<\\alpha<1$ provided that $\\Delta w\\in L^n(\\IB)$.\n",
"title": "Quasiconformal mappings and Hölder continuity"
}
| null | null | null | null | true | null |
14229
| null |
Default
| null | null |
null |
{
"abstract": " We prove that averaging operators are uniformly bounded on $L^1$ for all\ngeometrically doubling metric measure spaces, with bounds independent of the\nmeasure. From this result, the $L^1$ convergence of averages as $r \\to 0$\nimmediately follows.\n",
"title": "Boundedness of averaging operators on geometrically doubling metric spaces"
}
| null | null | null | null | true | null |
14230
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of computing first-passage time distributions for\nreaction processes modelled by master equations. We show that this generally\nintractable class of problems is equivalent to a sequential Bayesian inference\nproblem for an auxiliary observation process. The solution can be approximated\nefficiently by solving a closed set of coupled ordinary differential equations\n(for the low-order moments of the process) whose size scales with the number of\nspecies. We apply it to an epidemic model and a trimerisation process, and show\ngood agreement with stochastic simulations.\n",
"title": "Efficient Low-Order Approximation of First-Passage Time Distributions"
}
| null | null | null | null | true | null |
14231
| null |
Default
| null | null |
null |
{
"abstract": " Hamamatsu Photonics introduced a new generation of their Multi-Pixel Photon\nCounters in 2013 with significantly reduced after-pulsing rate. In this paper,\nwe investigate the causes of after-pulsing by testing pre-2013 and post-2013\ndevices using laser light ranging from 405 to 820nm. Doing so we investigate\nthe possibility that afterpulsing is also due to optical photons produced in\nthe avalanche rather than to impurities trapping charged carriers produced in\nthe avalanches and releasing them at a later time. For pre-2013 devices, we\nobserve avalanches delayed by ns to several 100~ns at 637, 777nm and 820 nm\ndemonstrating that holes created in the zero field region of the silicon bulk\ncan diffuse back to the high field region triggering delayed avalanches. On the\nother hand post-2013 exhibit no delayed avalanches beyond 100~ns at 777nm. We\nalso confirm that post-2013 devices exhibit about 25 times lower after-pulsing.\nTaken together, our measurements show that the absorption of photons from the\navalanche in the bulk of the silicon and the subsequent hole diffusion back to\nthe junction was a significant source of after-pulse for the pre-2013 devices.\nHamamatsu appears to have fixed this problem in 2013 following the preliminary\nrelease of our results. We also show that even at short wavelength the timing\ndistribution exhibit tails in the sub-nanosecond range that may impair the MPPC\ntiming performances.\n",
"title": "Delayed avalanches in Multi-Pixel Photon Counters"
}
| null | null | null | null | true | null |
14232
| null |
Default
| null | null |
null |
{
"abstract": " We introduce Dynamic Deep Neural Networks (D2NN), a new type of feed-forward\ndeep neural network that allows selective execution. Given an input, only a\nsubset of D2NN neurons are executed, and the particular subset is determined by\nthe D2NN itself. By pruning unnecessary computation depending on input, D2NNs\nprovide a way to improve computational efficiency. To achieve dynamic selective\nexecution, a D2NN augments a feed-forward deep neural network (directed acyclic\ngraph of differentiable modules) with controller modules. Each controller\nmodule is a sub-network whose output is a decision that controls whether other\nmodules can execute. A D2NN is trained end to end. Both regular and controller\nmodules in a D2NN are learnable and are jointly trained to optimize both\naccuracy and efficiency. Such training is achieved by integrating\nbackpropagation with reinforcement learning. With extensive experiments of\nvarious D2NN architectures on image classification tasks, we demonstrate that\nD2NNs are general and flexible, and can effectively optimize\naccuracy-efficiency trade-offs.\n",
"title": "Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Selective Execution"
}
| null | null | null | null | true | null |
14233
| null |
Default
| null | null |
null |
{
"abstract": " Gaussian random fields (GRF) are a fundamental stochastic model for\nspatiotemporal data analysis. An essential ingredient of GRF is the covariance\nfunction that characterizes the joint Gaussian distribution of the field.\nCommonly used covariance functions give rise to fully dense and unstructured\ncovariance matrices, for which required calculations are notoriously expensive\nto carry out for large data. In this work, we propose a construction of\ncovariance functions that result in matrices with a hierarchical structure.\nEmpowered by matrix algorithms that scale linearly with the matrix dimension,\nthe hierarchical structure is proved to be efficient for a variety of random\nfield computations, including sampling, kriging, and likelihood evaluation.\nSpecifically, with $n$ scattered sites, sampling and likelihood evaluation has\nan $O(n)$ cost and kriging has an $O(\\log n)$ cost after preprocessing,\nparticularly favorable for the kriging of an extremely large number of sites\n(e.g., predicting on more sites than observed). We demonstrate comprehensive\nnumerical experiments to show the use of the constructed covariance functions\nand their appealing computation time. Numerical examples on a laptop include\nsimulated data of size up to one million, as well as a climate data product\nwith over two million observations.\n",
"title": "Linear-Cost Covariance Functions for Gaussian Random Fields"
}
| null | null | null | null | true | null |
14234
| null |
Default
| null | null |
null |
{
"abstract": " The computability power of a distributed computing model is determined by the\ncommunication media available to the processes, the timing assumptions about\nprocesses and communication, and the nature of failures that processes can\nsuffer. In a companion paper we showed how dynamic epistemic logic can be used\nto give a formal semantics to a given distributed computing model, to capture\nprecisely the knowledge needed to solve a distributed task, such as consensus.\nFurthermore, by moving to a dual model of epistemic logic defined by simplicial\ncomplexes, topological invariants are exposed, which determine task\nsolvability. In this paper we show how to extend the setting above to include\nin the knowledge of the processes, knowledge about the model of computation\nitself. The extension describes the knowledge processes gain about the current\nexecution, in problems where processes have no input values at all.\n",
"title": "Models of fault-tolerant distributed computation via dynamic epistemic logic"
}
| null | null | null | null | true | null |
14235
| null |
Default
| null | null |
null |
{
"abstract": " This note corrects conditions in Proposition 3.4 and Theorem 5.2(ii) and\ncomments on imprecisions in Propositions 4.2 and 4.4 in Fissler and Ziegel\n(2016).\n",
"title": "Erratum: Higher Order Elicitability and Osband's Principle"
}
| null | null |
[
"Mathematics",
"Statistics",
"Quantitative Finance"
] | null | true | null |
14236
| null |
Validated
| null | null |
null |
{
"abstract": " We demonstrate an application of the Futamura Projections to human-computer\ninteraction, and particularly to staging human-computer dialogs. Specifically,\nby providing staging analogs to the classical Futamura Projections, we\ndemonstrate that the Futamura Projections can be applied to the staging of\nhuman-computer dialogs in addition to the execution of programs.\n",
"title": "Staging Human-computer Dialogs: An Application of the Futamura Projections"
}
| null | null | null | null | true | null |
14237
| null |
Default
| null | null |
null |
{
"abstract": " We report enhancing of complete synchronization in identical chaotic\noscillators when their interaction is mediated by a mismatched oscillator. The\nidentical oscillators now interact indirectly through the intermediate relay\noscillator. The induced heterogeneity in the intermediate oscillator plays a\nconstructive role in reducing the critical coupling for a transition to\ncomplete synchronization. A common lag synchronization emerges between the\nmismatched relay oscillator and its neighboring identical oscillators that\nleads to this enhancing effect. We present examples of one-dimensional open\narray, a ring, a star network and a two-dimensional lattice of dynamical\nsystems to demonstrate how this enhancing effect occurs. The paradigmatic\nRössler oscillator is used as a dynamical unit, in our numerical experiment,\nfor different networks to reveal the enhancing phenomenon.\n",
"title": "Enhancing synchronization in chaotic oscillators by induced heterogeneity"
}
| null | null | null | null | true | null |
14238
| null |
Default
| null | null |
null |
{
"abstract": " We propose a novel method for compressed sensing recovery using untrained\ndeep generative models. Our method is based on the recently proposed Deep Image\nPrior (DIP), wherein the convolutional weights of the network are optimized to\nmatch the observed measurements. We show that this approach can be applied to\nsolve any differentiable inverse problem. We also introduce a novel learned\nregularization technique which incorporates a small amount of prior\ninformation, further reducing the number of measurements required for a given\nreconstruction error. Our algorithm requires approximately 4-6x fewer\nmeasurements than classical Lasso methods. Unlike previous approaches based on\ngenerative models, our method does not require the model to be pre-trained. As\nsuch, we can apply our method to various medical imaging datasets for which\ndata acquisition is expensive and no known generative models exist.\n",
"title": "Compressed Sensing with Deep Image Prior and Learned Regularization"
}
| null | null | null | null | true | null |
14239
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, an energy harvesting scheme for a multi-user\nmultiple-input-multiple-output (MIMO) secrecy channel with artificial noise\n(AN) transmission is investigated. Joint optimization of the transmit\nbeamforming matrix, the AN covariance matrix, and the power splitting ratio is\nconducted to minimize the transmit power under the target secrecy rate, the\ntotal transmit power, and the harvested energy constraints. The original\nproblem is shown to be non-convex, which is tackled by a two-layer\ndecomposition approach. The inner layer problem is solved through semi-definite\nrelaxation, and the outer problem, on the other hand, is shown to be a single-\nvariable optimization that can be solved by one-dimensional (1- D) line search.\nTo reduce computational complexity, a sequential parametric convex\napproximation (SPCA) method is proposed to find a near-optimal solution. The\nwork is then extended to the imperfect channel state information case with\nnorm-bounded channel errors. Furthermore, tightness of the relaxation for the\nproposed schemes are validated by showing that the optimal solution of the\nrelaxed problem is rank-one. Simulation results demonstrate that the proposed\nSPCA method achieves the same performance as the scheme based on 1-D but with\nmuch lower complexity.\n",
"title": "Beamforming and Power Splitting Designs for AN-aided Secure Multi-user MIMO SWIPT Systems"
}
| null | null | null | null | true | null |
14240
| null |
Default
| null | null |
null |
{
"abstract": " We investigate spectral properties of the tensor products of two quantum\nchannels defined on matrix algebras. This leads to the important question of\nwhen an arbitrary subalgebra can split into the tensor product of two\nsubalgebras. We show that for two unital quantum channels $\\mathcal{E}_1$ and\n$\\mathcal{E}_2$ the multiplicative domain of\n$\\mathcal{E}_1\\otimes\\mathcal{E}_2$ splits into the tensor product of the\nindividual multiplicative domains. Consequently, we fully describe the fixed\npoints and peripheral eigen operators of the tensor product of channels.\nThrough a structure theorem of maximal unital proper $^*$-subalgebras (MUPSA)\nof a matrix algebra we provide a non-trivial upper bound of the 'multiplicative\nindex' of a unital channel which was recently introduced. This bound gives a\ncriteria on when a channel cannot be factored into a product of two different\nchannels. We construct examples of channels which can not be realized as a\ntensor product of two channels in any way. With these techniques and results,\nwe found some applications in quantum error correction.\n",
"title": "Spectral Properties of Tensor Products of Channels"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14241
| null |
Validated
| null | null |
null |
{
"abstract": " We prove that if a smooth projective algebraic variety of dimension less or\nequal to three has a unit type integral $K$-motive, then its integral Chow\nmotive is of Lefschetz type. As a consequence, the integral Chow motive is of\nLefschetz type for a smooth projective variety of dimension less or equal to\nthree that admits a full exceptional collection.\n",
"title": "Integral Chow motives of threefolds with $K$-motives of unit type"
}
| null | null | null | null | true | null |
14242
| null |
Default
| null | null |
null |
{
"abstract": " Recurrent Neural Networks (RNNs) are becoming increasingly important for time\nseries-related applications which require efficient and real-time\nimplementations. The recent pruning based work ESE suffers from degradation of\nperformance/energy efficiency due to the irregular network structure after\npruning. We propose block-circulant matrices for weight matrix representation\nin RNNs, thereby achieving simultaneous model compression and acceleration. We\naim to implement RNNs in FPGA with highest performance and energy efficiency,\nwith certain accuracy requirement (negligible accuracy degradation).\nExperimental results on actual FPGA deployments shows that the proposed\nframework achieves a maximum energy efficiency improvement of 35.7$\\times$\ncompared with ESE.\n",
"title": "Efficient Recurrent Neural Networks using Structured Matrices in FPGAs"
}
| null | null | null | null | true | null |
14243
| null |
Default
| null | null |
null |
{
"abstract": " Deep Learning has enabled remarkable progress over the last years on a\nvariety of tasks, such as image recognition, speech recognition, and machine\ntranslation. One crucial aspect for this progress are novel neural\narchitectures. Currently employed architectures have mostly been developed\nmanually by human experts, which is a time-consuming and error-prone process.\nBecause of this, there is growing interest in automated neural architecture\nsearch methods. We provide an overview of existing work in this field of\nresearch and categorize them according to three dimensions: search space,\nsearch strategy, and performance estimation strategy.\n",
"title": "Neural Architecture Search: A Survey"
}
| null | null |
[
"Statistics"
] | null | true | null |
14244
| null |
Validated
| null | null |
null |
{
"abstract": " With the advancement of technology in the last few decades, leading to the\nwidespread availability of miniaturized sensors and internet-connected things\n(IoT), security of electronic devices has become a top priority. Side-channel\nattack (SCA) is one of the prominent methods to break the security of an\nencryption system by exploiting the information leaked from the physical\ndevices. Correlational power attack (CPA) is an efficient power side-channel\nattack technique, which analyses the correlation between the estimated and\nmeasured supply current traces to extract the secret key. The existing\ncountermeasures to the power attacks are mainly based on reducing the SNR of\nthe leaked data, or introducing large overhead using techniques like power\nbalancing. This paper presents an attenuated signature AES (AS-AES), which\nresists SCA with minimal noise current overhead. AS-AES uses a shunt\nlow-drop-out (LDO) regulator to suppress the AES current signature by 400x in\nthe supply current traces. The shunt LDO has been fabricated and validated in\n130 nm CMOS technology. System-level implementation of the AS-AES along with\nnoise injection, shows that the system remains secure even after 50K\nencryptions, with 10x reduction in power overhead compared to that of noise\naddition alone.\n",
"title": "High Efficiency Power Side-Channel Attack Immunity using Noise Injection in Attenuated Signature Domain"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14245
| null |
Validated
| null | null |
null |
{
"abstract": " In this work thin magnetite films were deposited on SrTiO$_3$ via reactive\nmolecular beam epitaxy at different substrate temperatures. The growth process\nwas monitored in-situ during deposition by means of x-ray diffraction. While\nthe magnetite film grown at 400$^\\circ$C shows a fully relaxed vertical lattice\nconstant already in the early growth stages, the film deposited at 270$^\\circ$C\nexhibits a strong vertical compressive strain and relaxes towards the bulk\nvalue with increasing film thickness. Furthermore, a lateral tensile strain was\nobserved under these growth conditions although the inverse behavior is\nexpected due to the lattice mismatch of -7.5%. Additionally, the occupancy of\nthe A and B sublattices of magnetite with tetrahedral and octahedral sites was\ninvestigated showing a lower occupancy of the A sites compared to an ideal\ninverse spinel structure. The occupation of A sites decreases for a higher\ngrowth temperature. Thus, we assume a relocation of the iron ions from\ntetrahedral sites to octahedral vacancies forming a deficient rock salt\nlattice.\n",
"title": "Real-time monitoring of the structure of ultra thin Fe$_3$O$_4$ films during growth on Nb-doped SrTiO$_3$(001)"
}
| null | null | null | null | true | null |
14246
| null |
Default
| null | null |
null |
{
"abstract": " Demand response (DR) is a cost-effective and environmentally friendly\napproach for mitigating the uncertainties in renewable energy integration by\ntaking advantage of the flexibility of customers' demands. However, existing DR\nprograms suffer from either low participation due to strict commitment\nrequirements or not being reliable in voluntary programs. In addition, the\ncapacity planning for energy storage/reserves is traditionally done separately\nfrom the demand response program design, which incurs inefficiencies. Moreover,\ncustomers often face high uncertainties in their costs in providing demand\nresponse, which is not well studied in literature.\nThis paper first models the problem of joint capacity planning and demand\nresponse program design by a stochastic optimization problem, which\nincorporates the uncertainties from renewable energy generation, customer power\ndemands, as well as the customers' costs in providing DR. We propose online DR\ncontrol policies based on the optimal structures of the offline solution. A\ndistributed algorithm is then developed for implementing the control policies\nwithout efficiency loss. We further offer enhanced policy design by allowing\nflexibilities into the commitment level. We perform real world trace based\nnumerical simulations. Results demonstrate that the proposed algorithms can\nachieve near optimal social costs, and significant social cost savings compared\nto baseline methods.\n",
"title": "Harnessing Flexible and Reliable Demand Response Under Customer Uncertainties"
}
| null | null | null | null | true | null |
14247
| null |
Default
| null | null |
null |
{
"abstract": " The goal of graph representation learning is to embed each vertex in a graph\ninto a low-dimensional vector space. Existing graph representation learning\nmethods can be classified into two categories: generative models that learn the\nunderlying connectivity distribution in the graph, and discriminative models\nthat predict the probability of edge existence between a pair of vertices. In\nthis paper, we propose GraphGAN, an innovative graph representation learning\nframework unifying above two classes of methods, in which the generative model\nand discriminative model play a game-theoretical minimax game. Specifically,\nfor a given vertex, the generative model tries to fit its underlying true\nconnectivity distribution over all other vertices and produces \"fake\" samples\nto fool the discriminative model, while the discriminative model tries to\ndetect whether the sampled vertex is from ground truth or generated by the\ngenerative model. With the competition between these two models, both of them\ncan alternately and iteratively boost their performance. Moreover, when\nconsidering the implementation of generative model, we propose a novel graph\nsoftmax to overcome the limitations of traditional softmax function, which can\nbe proven satisfying desirable properties of normalization, graph structure\nawareness, and computational efficiency. Through extensive experiments on\nreal-world datasets, we demonstrate that GraphGAN achieves substantial gains in\na variety of applications, including link prediction, node classification, and\nrecommendation, over state-of-the-art baselines.\n",
"title": "GraphGAN: Graph Representation Learning with Generative Adversarial Nets"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14248
| null |
Validated
| null | null |
null |
{
"abstract": " There is widespread sentiment that it is not possible to effectively utilize\nfast gradient methods (e.g. Nesterov's acceleration, conjugate gradient, heavy\nball) for the purposes of stochastic optimization due to their instability and\nerror accumulation, a notion made precise in d'Aspremont 2008 and Devolder,\nGlineur, and Nesterov 2014. This work considers these issues for the special\ncase of stochastic approximation for the least squares regression problem, and\nour main result refutes the conventional wisdom by showing that acceleration\ncan be made robust to statistical errors. In particular, this work introduces\nan accelerated stochastic gradient method that provably achieves the minimax\noptimal statistical risk faster than stochastic gradient descent. Critical to\nthe analysis is a sharp characterization of accelerated stochastic gradient\ndescent as a stochastic process. We hope this characterization gives insights\ntowards the broader question of designing simple and effective accelerated\nstochastic methods for more general convex and non-convex optimization\nproblems.\n",
"title": "Accelerating Stochastic Gradient Descent For Least Squares Regression"
}
| null | null | null | null | true | null |
14249
| null |
Default
| null | null |
null |
{
"abstract": " We study, by means of the density-matrix renormalization group (DMRG)\ntechnique, the evolution of the ground state in a one-dimensional topological\ninsulator, from the non-interacting to the strongly-interacting limit, where\nthe system can be mapped onto a topological Kondo-insulator model. We focus on\na toy model Hamiltonian (i.e., the interacting \"$sp$-ladder\" model), which\ncould be experimentally realized in optical lattices with higher orbitals\nloaded with ultra-cold fermionic atoms. Our goal is to shed light on the\nemergence of the strongly-interacting ground state and its topological\nclassification as the Hubbard-$U$ interaction parameter of the model is\nincreased. Our numerical results show that the ground state can be generically\nclassified as a symmetry-protected topological phase of the Haldane-type, even\nin the non-interacting case $U=0$ where the system can be additionally\nclassified as a time-reversal $\\mathbb{Z}_{2}$-topological insulator, and\nevolves adiabatically between the non-interacting and strongly interacting\nlimits.\n",
"title": "Topological Kondo insulators in one dimension: Continuous Haldane-type ground-state evolution from the strongly-interacting to the non-interacting limit"
}
| null | null | null | null | true | null |
14250
| null |
Default
| null | null |
null |
{
"abstract": " Application of fuzzy support vector machine in stock price forecast. Support\nvector machine is a new type of machine learning method proposed in 1990s. It\ncan deal with classification and regression problems very successfully. Due to\nthe excellent learning performance of support vector machine, the technology\nhas become a hot research topic in the field of machine learning, and it has\nbeen successfully applied in many fields. However, as a new technology, there\nare many limitations to support vector machines. There is a large amount of\nfuzzy information in the objective world. If the training of support vector\nmachine contains noise and fuzzy information, the performance of the support\nvector machine will become very weak and powerless. As the complexity of many\nfactors influence the stock price prediction, the prediction results of\ntraditional support vector machine cannot meet people with precision, this\nstudy improved the traditional support vector machine fuzzy prediction\nalgorithm is proposed to improve the new model precision. NASDAQ Stock Market,\nStandard & Poor's (S&P) Stock market are considered. Novel advanced- fuzzy\nsupport vector machine (NA-FSVM) is the proposed methodology.\n",
"title": "A novel improved fuzzy support vector machine based stock price trend forecast model"
}
| null | null | null | null | true | null |
14251
| null |
Default
| null | null |
null |
{
"abstract": " We review the essentials of the formalism of quantum mechanics based on a\ndeformed Heisenbeg algebra, leading to the existence of a minimal length scale.\nWe compute in this context, the energy spectra of the pseudoharmonic oscillator\nand Kratzer potentials by using a perturbative approach. We derive the\nmolecular constants, which characterize the vibration--rotation energy levels\nof diatomic molecules, and investigate the effect of the minimal length on each\nof these parameters for both potentials. We confront our result to experimental\ndata for the hydrogen molecule to estimate an order of magnitude of this\nfundamental scale in molecular physics.\n",
"title": "Deformed Heisenberg Algebra with a minimal length: Application to some molecular potentials"
}
| null | null | null | null | true | null |
14252
| null |
Default
| null | null |
null |
{
"abstract": " Chemical substitution during growth is a well-established method to\nmanipulate electronic states of quantum materials, and leads to rich spectra of\nphase diagrams in cuprate and iron-based superconductors. Here we report a\nnovel and generic strategy to achieve nonvolatile electron doping in series of\n(i.e. 11 and 122 structures) Fe-based superconductors by ionic liquid gating\ninduced protonation at room temperature. Accumulation of protons in bulk\ncompounds induces superconductivity in the parent compounds, and enhances the\nTc largely in some superconducting ones. Furthermore, the existence of proton\nin the lattice enables the first proton nuclear magnetic resonance (NMR) study\nto probe directly superconductivity. Using FeS as a model system, our NMR study\nreveals an emergent high-Tc phase with no coherence peak which is hard to\nmeasure by NMR with other isotopes. This novel electric-field-induced proton\nevolution opens up an avenue for manipulation of competing electronic states\n(e.g. Mott insulators), and may provide an innovative way for a broad\nperspective of NMR measurements with greatly enhanced detecting resolution.\n",
"title": "Protonation induced high-Tc phases in iron-based superconductors evidenced by NMR and magnetization measurements"
}
| null | null | null | null | true | null |
14253
| null |
Default
| null | null |
null |
{
"abstract": " The relative root mean squared errors (RMSE) of nonparametric methods for\nspectral estimation is compared for microwave scattering data of plasma\nfluctuations. These methods reduce the variance of the periodogram estimate by\naveraging the spectrum over a frequency bandwidth. As the bandwidth increases,\nthe variance decreases, but the bias error increases. The plasma spectra vary\nby over four orders of magnitude, and therefore, using a spectral window is\nnecessary. We compare the smoothed tapered periodogram with the adaptive\nmultiple taper methods and hybrid methods. We find that a hybrid method, which\nuses four orthogonal tapers and then applies a kernel smoother, performs best.\nFor 300 point data segments, even an optimized smoothed tapered periodogram has\na 24 \\% larger relative RMSE than the hybrid method. We present two new\nadaptive multi-taper weightings which outperform Thomson's original adaptive\nweighting.\n",
"title": "Spectral Estimation of Plasma Fluctuations I: Comparison of Methods"
}
| null | null | null | null | true | null |
14254
| null |
Default
| null | null |
null |
{
"abstract": " The goal of the paper is to study the angle between two curves in the\nframework of metric (and metric measure) spaces. More precisely, we give a new\nnotion of angle between two curves in a metric space. Such a notion has a\nnatural interplay with optimal transportation and is particularly well suited\nfor metric measure spaces satisfying the curvature-dimension condition. Indeed\none of the main results is the validity of the cosine formula on $RCD^{*}(K,N)$\nmetric measure spaces. As a consequence, the new introduced notions are\ncompatible with the corresponding classical ones for Riemannian manifolds,\nRicci limit spaces and Alexandrov spaces.\n",
"title": "Angles between curves in metric measure spaces"
}
| null | null | null | null | true | null |
14255
| null |
Default
| null | null |
null |
{
"abstract": " The ability to locally degrade the extracellular matrix (ECM) and interact\nwith the tumour microenvironment is a key process distinguishing cancer from\nnormal cells, and is a critical step in the metastatic spread of the tumour.\nThe invasion of the surrounding tissue involves the coordinated action between\ncancer cells, the ECM, the matrix degrading enzymes, and the\nepithelial-to-mesenchymal transition (EMT). This is a regulatory process\nthrough which epithelial cells (ECs) acquire mesenchymal characteristics and\ntransform to mesenchymal-like cells (MCs). In this paper, we present a new\nmathematical model which describes the transition from a collective invasion\nstrategy for the ECs to an individual invasion strategy for the MCs. We achieve\nthis by formulating a coupled hybrid system consisting of partial and\nstochastic differential equations that describe the evolution of the ECs and\nthe MCs, respectively. This approach allows one to reproduce in a very natural\nway fundamental qualitative features of the current biomedical understanding of\ncancer invasion that are not easily captured by classical modelling approaches,\nfor example, the invasion of the ECM by self-generated gradients and the\nappearance of EC invasion islands outside of the main body of the tumour.\n",
"title": "A Hybrid Multiscale Model for Cancer Invasion of the Extracellular Matrix"
}
| null | null | null | null | true | null |
14256
| null |
Default
| null | null |
null |
{
"abstract": " In mathematical physics, the space-fractional diffusion equations are of\nparticular interest in the studies of physical phenomena modelled by Lévy\nprocesses, which are sometimes called super-diffusion equations. In this\narticle, we develop the differential quadrature (DQ) methods for solving the 2D\nspace-fractional diffusion equations on irregular domains. The methods in\npresence reduce the original equation into a set of ordinary differential\nequations (ODEs) by introducing valid DQ formulations to fractional directional\nderivatives based on the functional values at scattered nodal points on problem\ndomain. The required weighted coefficients are calculated by using radial basis\nfunctions (RBFs) as trial functions, and the resultant ODEs are discretized by\nthe Crank-Nicolson scheme. The main advantages of our methods lie in their\nflexibility and applicability to arbitrary domains. A series of illustrated\nexamples are finally provided to support these points.\n",
"title": "Differential quadrature method for space-fractional diffusion equations on 2D irregular domains"
}
| null | null | null | null | true | null |
14257
| null |
Default
| null | null |
null |
{
"abstract": " We derive some Positivstellensatzë for noncommutative rational expressions\nfrom the Positivstellensatzë for noncommutative polynomials. Specifically, we\nshow that if a noncommutative rational expression is positive on a polynomially\nconvex set, then there is an algebraic certificate witnessing that fact. As in\nthe case of noncommutative polynomials, our results are nicer when we\nadditionally assume positivity on a convex set-- that is, we obtain a so-called\n\"perfect Positivstellensatz\" on convex sets.\n",
"title": "Positivstellensatzë for noncommutative rational expressions"
}
| null | null | null | null | true | null |
14258
| null |
Default
| null | null |
null |
{
"abstract": " In the wake of the vast population of smart device users worldwide, mobile\nhealth (mHealth) technologies are hopeful to generate positive and wide\ninfluence on people's health. They are able to provide flexible, affordable and\nportable health guides to device users. Current online decision-making methods\nfor mHealth assume that the users are completely heterogeneous. They share no\ninformation among users and learn a separate policy for each user. However,\ndata for each user is very limited in size to support the separate online\nlearning, leading to unstable policies that contain lots of variances. Besides,\nwe find the truth that a user may be similar with some, but not all, users, and\nconnected users tend to have similar behaviors. In this paper, we propose a\nnetwork cohesion constrained (actor-critic) Reinforcement Learning (RL) method\nfor mHealth. The goal is to explore how to share information among similar\nusers to better convert the limited user information into sharper learned\npolicies. To the best of our knowledge, this is the first online actor-critic\nRL for mHealth and first network cohesion constrained (actor-critic) RL method\nin all applications. The network cohesion is important to derive effective\npolicies. We come up with a novel method to learn the network by using the warm\nstart trajectory, which directly reflects the users' property. The optimization\nof our model is difficult and very different from the general supervised\nlearning due to the indirect observation of values. As a contribution, we\npropose two algorithms for the proposed online RLs. Apart from mHealth, the\nproposed methods can be easily applied or adapted to other health-related\ntasks. Extensive experiment results on the HeartSteps dataset demonstrates that\nin a variety of parameter settings, the proposed two methods obtain obvious\nimprovements over the state-of-the-art methods.\n",
"title": "Cohesion-based Online Actor-Critic Reinforcement Learning for mHealth Intervention"
}
| null | null | null | null | true | null |
14259
| null |
Default
| null | null |
null |
{
"abstract": " We study the minus order on the algebra of bounded linear operators on a\nHilbert space. By giving a characterization in terms of range additivity, we\nshow that the intrinsic nature of the minus order is algebraic. Applications to\ngeneralized inverses of the sum of two operators, to systems of operator\nequations and to optimization problems are also presented.\n",
"title": "The minus order and range additivity"
}
| null | null | null | null | true | null |
14260
| null |
Default
| null | null |
null |
{
"abstract": " While the dynamics of a fully flexible polymer ejecting a capsid through a\nnanopore has been extensively studied, the ejection dynamics of semiflexible\npolymers has not been properly characterized. Here we report results from\nsimulations of ejection dynamics of semiflexible polymers ejecting from\nspherical capsids. Ejections start from strongly confined polymer conformations\nof constant initial monomer density. We find that, unlike for fully flexible\npolymers, for semiflexible polymers the force measured at the pore does not\nshow a direct relation to the instantaneous ejection velocity. The cumulative\nwaiting time $t(s)$, that is, the time at which a monomer $s$ exits the capsid\nthe last time, shows a clear change when increasing the polymer rigidity\n$\\kappa$. Major part of an ejecting polymer is driven out of the capsid by\ninternal pressure. At the final stage the polymer escapes the capsid by\ndiffusion. For the driven part there is a cross-over from essentially\nexponential growth of $t$ with $s$ of the fully flexible polymers to a\nscale-invariant form. In addition, a clear dependence of $t$ on $N_0$ was\nfound. These findings combined give the dependence $t(s) \\propto N_0^{0.55}\ns^{1.33}$ for the strongly rigid polymers. This cross-over in dynamics where\n$\\kappa$ acts as a control parameter is reminiscent of a phase transition. This\nanalogy is further enhanced by our finding a perfect data collapse of $t$ for\npolymers of different $N_0$ and any constant $\\kappa$.\n",
"title": "Rigidity-induced scale invariance in polymer ejection from capsid"
}
| null | null | null | null | true | null |
14261
| null |
Default
| null | null |
null |
{
"abstract": " We have created a cloud-based service that allows the end users to run tests\non multiple different databases to find which databases are most suitable for\ntheir project. From our research, we could not find another application that\nenables the user to test several databases to gauge the difference between\nthem. This application allows the user to choose which type of test to perform\nand which databases to target. The application also displays the results of\ndifferent tests that were run by other users previously. There is also a map to\nshow the location where all the tests are run to give the user an estimate of\nthe location. Unlike the orthodox static tests and reports conducted to\nevaluate NoSQL databases, we have created a web application to run and analyze\nthese tests in real time. This web application evaluates the performance of\nseveral NoSQL databases. The databases covered are MongoDB, DynamoDB, CouchDB,\nand Firebase. The web service is accessible from: nosqldb.nextproject.ca.\n",
"title": "A Cloud-based Service for Real-Time Performance Evaluation of NoSQL Databases"
}
| null | null | null | null | true | null |
14262
| null |
Default
| null | null |
null |
{
"abstract": " A ballean (or coarse structure) is a set endowed with some family of subsets,\nthe balls, is such a way that balleans with corresponding morphisms can be\nconsidered as asymptotic counterparts of uniform topological spaces. For a\nballean $\\mathcal{B}$ on a set $X$, the hyperballean $\\mathcal{B}^{\\flat}$ is a\nballean naturally defined on the set $X^{\\flat}$ of all bounded subsets of $X$.\nWe describe all balleans with hyperballeans of bounded geometry and analyze the\nstructure of these hyperballeans.\n",
"title": "On hyperballeans of bounded geometry"
}
| null | null | null | null | true | null |
14263
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we investigate large sample properties of model selection\nprocedures in a general Bayesian framework when a closed form expression of the\nmarginal likelihood function is not available or a local asymptotic quadratic\napproximation of the log-likelihood function does not exist. Under appropriate\nidentifiability assumptions on the true model, we provide sufficient conditions\nfor a Bayesian model selection procedure to be consistent and obey the Occam's\nrazor phenomenon, i.e., the probability of selecting the \"smallest\" model that\ncontains the truth tends to one as the sample size goes to infinity. In order\nto show that a Bayesian model selection procedure selects the smallest model\ncontaining the truth, we impose a prior anti-concentration condition, requiring\nthe prior mass assigned by large models to a neighborhood of the truth to be\nsufficiently small. In a more general setting where the strong model\nidentifiability assumption may not hold, we introduce the notion of local\nBayesian complexity and develop oracle inequalities for Bayesian model\nselection procedures. Our Bayesian oracle inequality characterizes a trade-off\nbetween the approximation error and a Bayesian characterization of the local\ncomplexity of the model, illustrating the adaptive nature of averaging-based\nBayesian procedures towards achieving an optimal rate of posterior convergence.\nSpecific applications of the model selection theory are discussed in the\ncontext of high-dimensional nonparametric regression and density regression\nwhere the regression function or the conditional density is assumed to depend\non a fixed subset of predictors. As a result of independent interest, we\npropose a general technique for obtaining upper bounds of certain small ball\nprobability of stationary Gaussian processes.\n",
"title": "Bayesian model selection consistency and oracle inequality with intractable marginal likelihood"
}
| null | null | null | null | true | null |
14264
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of generating relevant execution traces to test rich\ninteractive applications. Rich interactive applications, such as apps on mobile\nplatforms, are complex stateful and often distributed systems where\nsufficiently exercising the app with user-interaction (UI) event sequences to\nexpose defects is both hard and time-consuming. In particular, there is a\nfundamental tension between brute-force random UI exercising tools, which are\nfully-automated but offer low relevance, and UI test scripts, which are manual\nbut offer high relevance. In this paper, we consider a middle way---enabling a\nseamless fusion of scripted and randomized UI testing. This fusion is\nprototyped in a testing tool called ChimpCheck for programming, generating, and\nexecuting property-based randomized test cases for Android apps. Our approach\nrealizes this fusion by offering a high-level, embedded domain-specific\nlanguage for defining custom generators of simulated user-interaction event\nsequences. What follows is a combinator library built on industrial strength\nframeworks for property-based testing (ScalaCheck) and Android testing (Android\nJUnit and Espresso) to implement property-based randomized testing for Android\ndevelopment. Driven by real, reported issues in open source Android apps, we\nshow, through case studies, how ChimpCheck enables expressing effective testing\npatterns in a compact manner.\n",
"title": "ChimpCheck: Property-Based Randomized Test Generation for Interactive Apps"
}
| null | null | null | null | true | null |
14265
| null |
Default
| null | null |
null |
{
"abstract": " The dynamics along the particle trajectories for the 3D axisymmetric Euler\nequations are considered. It is shown that if the inflow is rapidly increasing\n(pushy) in time, the corresponding laminar profile of the incompressible Euler\nflow is not (in some sense) stable provided that the swirling component is not\nzero. It is also shown that if the vorticity on the axis is not zero (with some\nextra assumptions), then there is no steady flow. We can rephrase these\ninstability to an instantaneous blow-up. In the proof, Frenet-Serret formulas\nand orthonormal moving frame are essentially used.\n",
"title": "Mathematical analysis of pulsatile flow, vortex breakdown and instantaneous blow-up for the axisymmetric Euler equations"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14266
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, a resilient controller is designed for the linear\ntime-invariant (LTI) systems subject to attacks on the sensors and the\nactuators. A novel probabilistic attack model is proposed to capture\nvulnerabilities of the communication links from sensors to the controller and\nfrom the controller to actuators. The observer and the controller formulation\nunder the attack are derived. Thereafter, By leveraging Lyapunov functional\nmethods, it is shown that exponential mean-square stability of the system under\nthe output feedback controller is guaranteed if a certain LMI is feasible. The\nsimulation results show the effectiveness and applicability of the proposed\ncontroller design approach.\n",
"title": "Resilient Feedback Controller Design For Linear Model of Power Grids"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14267
| null |
Validated
| null | null |
null |
{
"abstract": " We prove that integrability of a dispersionless Hirota type equation implies\nthe symplectic Monge-Ampere property in any dimension $\\geq 4$. In 4D this\nyields a complete classification of integrable dispersionless PDEs of Hirota\ntype through a list of heavenly type equations arising in self-dual gravity. As\na by-product of our approach we derive an involutive system of relations\ncharacterising symplectic Monge-Ampere equations in any dimension.\nMoreover, we demonstrate that in 4D the requirement of integrability is\nequivalent to self-duality of the conformal structure defined by the\ncharacteristic variety of the equation on every solution, which is in turn\nequivalent to the existence of a dispersionless Lax pair. We also give a\ncriterion of linerisability of a Hirota type equation via flatness of the\ncorresponding conformal structure, and study symmetry properties of integrable\nequations.\n",
"title": "Integrability of dispersionless Hirota type equations in 4D and the symplectic Monge-Ampere property"
}
| null | null | null | null | true | null |
14268
| null |
Default
| null | null |
null |
{
"abstract": " In this note we propose a new approach towards solving numerically optimal\nstopping problems via reinforced regression based Monte Carlo algorithms. The\nmain idea of the method is to reinforce standard linear regression algorithms\nin each backward induction step by adding new basis functions based on\npreviously estimated continuation values. The proposed methodology is\nillustrated by a numerical example from mathematical finance.\n",
"title": "Optimal stopping via reinforced regression"
}
| null | null | null | null | true | null |
14269
| null |
Default
| null | null |
null |
{
"abstract": " We use the Maximum $q$-log-likelihood estimation for Least informative\ndistributions (LID) in order to estimate the parameters in probability density\nfunctions (PDFs) efficiently and robustly when data include outlier(s). LIDs\nare derived by using convex combinations of two PDFs,\n$f_\\epsilon=(1-\\epsilon)f_0+\\epsilon f_1$. A convex combination of two PDFs is\nconsidered as a contamination $f_1$ as outlier(s) to underlying $f_0$\ndistributions and $f_\\epsilon$ is a contaminated distribution. The optimal\ncriterion is obtained by minimizing the change of Maximum q-log-likelihood\nfunction when the data have slightly more contamination. In this paper, we make\na comparison among ordinary Maximum likelihood, Maximum q-likelihood\nestimations, LIDs based on $\\log_q$ and Huber M-estimation. Akaike and Bayesian\ninformation criterions (AIC and BIC) based on $\\log_q$ and LID are proposed to\nassess the fitting performance of functions. Real data sets are applied to test\nthe fitting performance of estimating functions that include shape, scale and\nlocation parameters.\n",
"title": "Least informative distributions in Maximum q-log-likelihood estimation"
}
| null | null | null | null | true | null |
14270
| null |
Default
| null | null |
null |
{
"abstract": " Groups of enterprises guarantee each other and form complex guarantee\nnetworks when they try to obtain loans from banks. Such secured loan can\nenhance the solvency and promote the rapid growth in the economic upturn\nperiod. However, potential systemic risk may happen within the risk binding\ncommunity. Especially, during the economic down period, the crisis may spread\nin the guarantee network like a domino. Monitoring the financial status,\npreventing or reducing systematic risk when crisis happens is highly concerned\nby the regulatory commission and banks. We propose visual analytics approach\nfor loan guarantee network risk management, and consolidate the five analysis\ntasks with financial experts: i) visual analytics for enterprises default risk,\nwhereby a hybrid representation is devised to predict the default risk and\ndeveloped an interface to visualize key indicators; ii) visual analytics for\nhigh default groups, whereby a community detection based interactive approach\nis presented; iii) visual analytics for high defaults pattern, whereby a motif\ndetection based interactive approach is described, and we adopt a Shneiderman\nMantra strategy to reduce the computation complexity. iv) visual analytics for\nevolving guarantee network, whereby animation is used to help understanding the\nguarantee dynamic; v) visual analytics approach and interface for default\ndiffusion path. The temporal diffusion path analysis can be useful for the\ngovernment and bank to monitor the default spread status. It also provides\ninsight for taking precautionary measures to prevent and dissolve systemic\nfinancial risk. We implement the system with case studies on a real-world\nguarantee network. Two financial experts are consulted with endorsement on the\ndeveloped tool. To the best of our knowledge, this is the first visual\nanalytics tool to explore the guarantee network risks in a systematic manner.\n",
"title": "Visual analytics for loan guarantee network risk management"
}
| null | null | null | null | true | null |
14271
| null |
Default
| null | null |
null |
{
"abstract": " The Vocal Joystick Vowel Corpus, by Washington University, was used to study\nmonophthongs pronounced by native English speakers. The objective of this study\nwas to quantitatively measure the extent at which speech recognition methods\ncan distinguish between similar sounding vowels. In particular, the phonemes\n/\\textipa{@}/, /{\\ae}/, /\\textipa{A}:/ and /\\textipa{2}/ were analysed. 748\nsound files from the corpus were used and subjected to Linear Predictive Coding\n(LPC) to compute their formants, and to Mel Frequency Cepstral Coefficients\n(MFCC) algorithm, to compute the cepstral coefficients. A Decision Tree\nClassifier was used to build a predictive model that learnt the patterns of the\ntwo first formants measured in the data set, as well as the patterns of the 13\ncepstral coefficients. An accuracy of 70\\% was achieved using formants for the\nmentioned phonemes. For the MFCC analysis an accuracy of 52 \\% was achieved and\nan accuracy of 71\\% when /\\textipa{@}/ was ignored. The results obtained show\nthat the studied algorithms are far from mimicking the ability of\ndistinguishing subtle differences in sounds like human hearing does.\n",
"title": "Pronunciation recognition of English phonemes /\\textipa{@}/, /æ/, /\\textipa{A}:/ and /\\textipa{2}/ using Formants and Mel Frequency Cepstral Coefficients"
}
| null | null | null | null | true | null |
14272
| null |
Default
| null | null |
null |
{
"abstract": " We examine salient trends of influenza pandemics in Australia, a rapidly\nurbanizing nation. To do so, we implement state-of-the-art influenza\ntransmission and progression models within a large-scale stochastic computer\nsimulation, generated using comprehensive Australian census datasets from 2006,\n2011, and 2016. Our results offer the first simulation-based investigation of a\npopulation's sensitivity to pandemics across multiple historical time points,\nand highlight three significant trends in pandemic patterns over the years:\nincreased peak prevalence, faster spreading rates, and decreasing\nspatiotemporal bimodality. We attribute these pandemic trends to increases in\ntwo key quantities indicative of urbanization: population fraction residing in\nmajor cities, and international air traffic. In addition, we identify features\nof the pandemic's geographic spread that can only be attributed to changes in\nthe commuter mobility network. The generic nature of our model and the ubiquity\nof urbanization trends around the world make it likely for our results to be\napplicable in other rapidly urbanizing nations.\n",
"title": "Vulnerability to pandemics in a rapidly urbanizing society"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
14273
| null |
Validated
| null | null |
null |
{
"abstract": " Henrik Bruus is professor of lab-chip systems and theoretical physics at the\nTechnical University of Denmark. In this contribution, he summarizes some of\nthe recent results within theory and simulation of microscale acoustofluidic\nsystems that he has obtained in collaboration with his students and\ninternational colleagues. The main emphasis is on three dynamical effects\ninduced by external ultrasound fields acting on aqueous solutions and particle\nsuspensions: The acoustic radiation force acting on suspended micro- and\nnanoparticles, the acoustic streaming appearing in the fluid, and the newly\ndiscovered acoustic body force acting on inhomogeneous solutions.\n",
"title": "Theoretical aspects of microscale acoustofluidics"
}
| null | null | null | null | true | null |
14274
| null |
Default
| null | null |
null |
{
"abstract": " Analysis tools like abstract interpreters, symbolic execution tools and\ntesting tools usually require a proper context to give useful results when\nanalyzing a particular function. Such a context initializes the function\nparameters and global variables to comply with function requirements. However\nit may be error-prone to write it by hand: the handwritten context might\ncontain bugs or not match the intended specification. A more robust approach is\nto specify the context in a dedicated specification language, and hold the\nanalysis tools to support it properly. This may mean to put significant\ndevelopment efforts for enhancing the tools, something that is often not\nfeasible if ever possible.\nThis paper presents a way to systematically generate such a context from a\nformal specification of a C function. This is applied to a subset of the ACSL\nspecification language in order to generate suitable contexts for the abstract\ninterpretation-based value analysis plug-ins of Frama-C, a framework for\nanalysis of code written in C. The idea here presented has been implemented in\na new Frama-C plug-in which is currently in use in an operational industrial\nsetting.\n",
"title": "Context Generation from Formal Specifications for C Analysis Tools"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14275
| null |
Validated
| null | null |
null |
{
"abstract": " This paper considers the problem of solving systems of quadratic equations,\nnamely, recovering an object of interest\n$\\mathbf{x}^{\\natural}\\in\\mathbb{R}^{n}$ from $m$ quadratic equations/samples\n$y_{i}=(\\mathbf{a}_{i}^{\\top}\\mathbf{x}^{\\natural})^{2}$, $1\\leq i\\leq m$. This\nproblem, also dubbed as phase retrieval, spans multiple domains including\nphysical sciences and machine learning.\nWe investigate the efficiency of gradient descent (or Wirtinger flow)\ndesigned for the nonconvex least squares problem. We prove that under Gaussian\ndesigns, gradient descent --- when randomly initialized --- yields an\n$\\epsilon$-accurate solution in $O\\big(\\log n+\\log(1/\\epsilon)\\big)$ iterations\ngiven nearly minimal samples, thus achieving near-optimal computational and\nsample complexities at once. This provides the first global convergence\nguarantee concerning vanilla gradient descent for phase retrieval, without the\nneed of (i) carefully-designed initialization, (ii) sample splitting, or (iii)\nsophisticated saddle-point escaping schemes. All of these are achieved by\nexploiting the statistical models in analyzing optimization algorithms, via a\nleave-one-out approach that enables the decoupling of certain statistical\ndependency between the gradient descent iterates and the data.\n",
"title": "Gradient Descent with Random Initialization: Fast Global Convergence for Nonconvex Phase Retrieval"
}
| null | null | null | null | true | null |
14276
| null |
Default
| null | null |
null |
{
"abstract": " Unwanted variation, including hidden confounding, is a well-known problem in\nmany fields, particularly large-scale gene expression studies. Recent proposals\nto use control genes --- genes assumed to be unassociated with the covariates\nof interest --- have led to new methods to deal with this problem. Going by the\nmoniker Removing Unwanted Variation (RUV), there are many versions --- RUV1,\nRUV2, RUV4, RUVinv, RUVrinv, RUVfun. In this paper, we introduce a general\nframework, RUV*, that both unites and generalizes these approaches. This\nunifying framework helps clarify connections between existing methods. In\nparticular we provide conditions under which RUV2 and RUV4 are equivalent. The\nRUV* framework also preserves an advantage of RUV approaches --- their\nmodularity --- which facilitates the development of novel methods based on\nexisting matrix imputation algorithms. We illustrate this by implementing RUVB,\na version of RUV* based on Bayesian factor analysis. In realistic simulations\nbased on real data we found that RUVB is competitive with existing methods in\nterms of both power and calibration, although we also highlight the challenges\nof providing consistently reliable calibration among data sets.\n",
"title": "Unifying and Generalizing Methods for Removing Unwanted Variation Based on Negative Controls"
}
| null | null | null | null | true | null |
14277
| null |
Default
| null | null |
null |
{
"abstract": " The purpose of this short contribution is to report on the development of a\nSpectral Neighbor Analysis Potential (SNAP) for tungsten. We have focused on\nthe characterization of elastic and defect properties of the pure material in\norder to support molecular dynamics simulations of plasma-facing materials in\nfusion reactors. A parallel genetic algorithm approach was used to efficiently\nsearch for fitting parameters optimized against a large number of objective\nfunctions. In addition, we have shown that this many-body tungsten potential\ncan be used in conjunction with a simple helium pair potential to produce\naccurate defect formation energies for the W-He binary system.\n",
"title": "Quantum-Accurate Molecular Dynamics Potential for Tungsten"
}
| null | null |
[
"Physics"
] | null | true | null |
14278
| null |
Validated
| null | null |
null |
{
"abstract": " Bandt and Pompe introduced Permutation Entropy in 2002 for Time Series where\nequal values, xt1 = xt2, t1 = t2, were neglected and only inequalities between\nthe xt were considered. Since then, this measure has been modified and\nextended, in particular in cases when the amount of equal values in the series\ncan not be neglected, (i.e. heart rate variability (HRV) time series). We\nreview the different existing methodologies that treats this subject by\nclassifying them according to their different strategies. In addition, a novel\nBayesian Missing Data Imputation is presented that proves to outperform the\nexisting methodologies that deals with type of time series. All this facts are\nillustrated by simulations and also by distinguishing patients suffering from\nCongestive Heart Failure from a (healthy) control group using HRV time series\n",
"title": "An empirical evaluation of alternative methods of estimation for Permutation Entropy in time series with tied values"
}
| null | null | null | null | true | null |
14279
| null |
Default
| null | null |
null |
{
"abstract": " Chest radiography is an extremely powerful imaging modality, allowing for a\ndetailed inspection of a patient's thorax, but requiring specialized training\nfor proper interpretation. With the advent of high performance general purpose\ncomputer vision algorithms, the accurate automated analysis of chest\nradiographs is becoming increasingly of interest to researchers. However, a key\nchallenge in the development of these techniques is the lack of sufficient\ndata. Here we describe MIMIC-CXR, a large dataset of 371,920 chest x-rays\nassociated with 227,943 imaging studies sourced from the Beth Israel Deaconess\nMedical Center between 2011 - 2016. Each imaging study can pertain to one or\nmore images, but most often are associated with two images: a frontal view and\na lateral view. Images are provided with 14 labels derived from a natural\nlanguage processing tool applied to the corresponding free-text radiology\nreports. All images have been de-identified to protect patient privacy. The\ndataset is made freely available to facilitate and encourage a wide range of\nresearch in medical computer vision.\n",
"title": "MIMIC-CXR: A large publicly available database of labeled chest radiographs"
}
| null | null | null | null | true | null |
14280
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents a Convolutional Neural Network (CNN) based page\nsegmentation method for handwritten historical document images. We consider\npage segmentation as a pixel labeling problem, i.e., each pixel is classified\nas one of the predefined classes. Traditional methods in this area rely on\ncarefully hand-crafted features or large amounts of prior knowledge. In\ncontrast, we propose to learn features from raw image pixels using a CNN. While\nmany researchers focus on developing deep CNN architectures to solve different\nproblems, we train a simple CNN with only one convolution layer. We show that\nthe simple architecture achieves competitive results against other deep\narchitectures on different public datasets. Experiments also demonstrate the\neffectiveness and superiority of the proposed method compared to previous\nmethods.\n",
"title": "Convolutional Neural Networks for Page Segmentation of Historical Document Images"
}
| null | null | null | null | true | null |
14281
| null |
Default
| null | null |
null |
{
"abstract": " Let $(x_n)_{n=1}^{\\infty}$ be a sequence on the torus $\\mathbb{T}$\n(normalized to length 1). We show that if there exists a sequence of positive\nreal numbers $(t_n)_{n=1}^{\\infty}$ converging to 0 such that $$ \\lim_{N\n\\rightarrow \\infty}{\\frac{1}{N^2} \\sum_{m,n = 1}^{N}{\\frac{1}{\\sqrt{t_N}}\n\\exp{\\left(- \\frac{1}{t_N} (x_m - x_n)^2 \\right)}}}\n= \\sqrt{\\pi},$$ then $(x_n)_{n=1}^{\\infty}$ is uniformly distributed. This is\nespecially interesting when $t_N \\sim N^{-2}$ since the size of the sum is then\nessentially determined exclusively by local gaps at scale $\\sim N^{-1}$. This\ncan be used to show equidistribution of sequences with Poissonian pair\ncorrelation, which recovers a recent result of Aistleitner, Lachmann &\nPausinger and Grepstad & Larcher. The general form of the result is proven on\narbitrary compact manifolds $(M,g)$ where the role of the exponential function\nis played by the heat kernel $e^{t\\Delta}$: for all $x_1, \\dots, x_N \\in M$ and\nall $t>0$ $$ \\frac{1}{N^2}\\sum_{m,n=1}^{N}{[e^{t\\Delta}\\delta_{x_m}](x_n)} \\geq\n\\frac{1}{vol(M)}$$ and equality is attained as $N \\rightarrow \\infty$ if and\nonly if $(x_n)_{n=1}^{\\infty}$ equidistributes.\n",
"title": "Localized Quantitative Criteria for Equidistribution"
}
| null | null | null | null | true | null |
14282
| null |
Default
| null | null |
null |
{
"abstract": " Conventional generators in power grids are steadily substituted with new\nrenewable sources of electric power. The latter are connected to the grid via\ninverters and as such have little, if any rotational inertia. The resulting\nreduction of total inertia raises important issues of power grid stability,\nespecially over short-time scales. We have constructed a model of the\nsynchronous grid of continental Europe with which we numerically investigate\nfrequency deviations as well as rates of change of frequency (RoCoF) following\nabrupt power losses. The magnitude of RoCoF's and frequency deviations strongly\ndepend on the fault location, and we find the largest effects for faults\nlocated on the slowest mode - the Fiedler mode - of the network Laplacian\nmatrix. This mode essentially vanishes over Belgium, Eastern France, Western\nGermany, northern Italy and Switzerland. Buses inside these regions are only\nweakly affected by faults occuring outside. Conversely, faults inside these\nregions have only a local effect and disturb only weakly outside buses.\nFollowing this observation, we reduce rotational inertia through three\ndifferent procedures by either (i) reducing inertia on the Fiedler mode, (ii)\nreducing inertia homogeneously and (iii) reducing inertia outside the Fiedler\nmode. We find that procedure (iii) has little effect on disturbance\npropagation, while procedure (i) leads to the strongest increase of RoCoF and\nfrequency deviations. These results for our model of the European transmission\ngrid are corroborated by numerical investigations on the ERCOT transmission\ngrid.\n",
"title": "Disturbance propagation, inertia location and slow modes in large-scale high voltage power grids"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14283
| null |
Validated
| null | null |
null |
{
"abstract": " One-sided matching mechanisms are fundamental for assigning a set of\nindivisible objects to a set of self-interested agents when monetary transfers\nare not allowed. Two widely-studied randomized mechanisms in multiagent\nsettings are the Random Serial Dictatorship (RSD) and the Probabilistic Serial\nRule (PS). Both mechanisms require only that agents specify ordinal preferences\nand have a number of desirable economic and computational properties. However,\nthe induced outcomes of the mechanisms are often incomparable and thus there\nare challenges when it comes to deciding which mechanism to adopt in practice.\nIn this paper, we first consider the space of general ordinal preferences and\nprovide empirical results on the (in)comparability of RSD and PS. We analyze\ntheir respective economic properties under general and lexicographic\npreferences. We then instantiate utility functions with the goal of gaining\ninsights on the manipulability, efficiency, and envyfreeness of the mechanisms\nunder different risk-attitude models. Our results hold under various preference\ndistribution models, which further confirm the broad use of RSD in most\npractical applications.\n",
"title": "Investigating the Characteristics of One-Sided Matching Mechanisms Under Various Preferences and Risk Attitudes"
}
| null | null | null | null | true | null |
14284
| null |
Default
| null | null |
null |
{
"abstract": " We introduce MIDI-VAE, a neural network model based on Variational\nAutoencoders that is capable of handling polyphonic music with multiple\ninstrument tracks, as well as modeling the dynamics of music by incorporating\nnote durations and velocities. We show that MIDI-VAE can perform style transfer\non symbolic music by automatically changing pitches, dynamics and instruments\nof a music piece from, e.g., a Classical to a Jazz style. We evaluate the\nefficacy of the style transfer by training separate style validation\nclassifiers. Our model can also interpolate between short pieces of music,\nproduce medleys and create mixtures of entire songs. The interpolations\nsmoothly change pitches, dynamics and instrumentation to create a harmonic\nbridge between two music pieces. To the best of our knowledge, this work\nrepresents the first successful attempt at applying neural style transfer to\ncomplete musical compositions.\n",
"title": "MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer"
}
| null | null | null | null | true | null |
14285
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we present Gumbel Graph Network, a model-free deep learning\nframework for dynamics learning and network reconstruction from the observed\ntime series data. Our method requires no prior knowledge about underlying\ndynamics and has shown the state-of-the-art performance in three typical\ndynamical systems on complex networks.\n",
"title": "A General Deep Learning Framework for Structure and Dynamics Reconstruction from Time Series Data"
}
| null | null | null | null | true | null |
14286
| null |
Default
| null | null |
null |
{
"abstract": " Majorana bound states (MBS) are well-established in the clean limit in chains\nof ferromagnetically aligned impurities deposited on conventional\nsuperconductors with finite spin-orbit coupling. Here we show that these MBS\nare very robust against disorder. By performing self-consistent calculations we\nfind that the MBS are protected as long as the surrounding superconductor show\nno large signs of inhomogeneity. We find that longer chains offer more\nstability against disorder for the MBS, albeit the minigap decreases, as do\nincreasing strengths of spin-orbit coupling and superconductivity.\n",
"title": "Disorder robustness and protection of Majorana bound states in ferromagnetic chains on conventional superconductors"
}
| null | null | null | null | true | null |
14287
| null |
Default
| null | null |
null |
{
"abstract": " An upgrade of the ATLAS experiment for the High Luminosity phase of LHC is\nplanned for 2024 and foresees the replacement of the present Inner Detector\n(ID) with a new Inner Tracker (ITk) completely made of silicon devices.\nDepleted active pixel sensors built with the High Voltage CMOS (HV-CMOS)\ntechnology are investigated as an option to cover large areas in the outermost\nlayers of the pixel detector and are especially interesting for the development\nof monolithic devices which will reduce the production costs and the material\nbudget with respect to the present hybrid assemblies. For this purpose the\nH35DEMO, a large area HV-CMOS demonstrator chip, was designed by KIT, IFAE and\nUniversity of Liverpool, and produced in AMS 350 nm CMOS technology. It\nconsists of four pixel matrices and additional test structures. Two of the\nmatrices include amplifiers and discriminator stages and are thus designed to\nbe operated as monolithic detectors. In these devices the signal is mainly\nproduced by charge drift in a small depleted volume obtained by applying a bias\nvoltage of the order of 100 V. Moreover, to enhance the radiation hardness of\nthe chip, this technology allows to enclose the electronics in the same deep\nN-WELLs which are also used as collecting electrodes. In this contribution the\ncharacterisation of H35DEMO chips and results of the very first beam test\nmeasurements of the monolithic CMOS matrices with high energetic pions at CERN\nSPS will be presented.\n",
"title": "Characterisation of novel prototypes of monolithic HV-CMOS pixel detectors for high energy physics experiments"
}
| null | null | null | null | true | null |
14288
| null |
Default
| null | null |
null |
{
"abstract": " Recently several end-to-end speaker verification systems based on deep neural\nnetworks (DNNs) have been proposed. These systems have been proven to be\ncompetitive for text-dependent tasks as well as for text-independent tasks with\nshort utterances. However, for text-independent tasks with longer utterances,\nend-to-end systems are still outperformed by standard i-vector + PLDA systems.\nIn this work, we develop an end-to-end speaker verification system that is\ninitialized to mimic an i-vector + PLDA baseline. The system is then further\ntrained in an end-to-end manner but regularized so that it does not deviate too\nfar from the initial system. In this way we mitigate overfitting which normally\nlimits the performance of end-to-end systems. The proposed system outperforms\nthe i-vector + PLDA baseline on both long and short duration utterances.\n",
"title": "End-to-end DNN Based Speaker Recognition Inspired by i-vector and PLDA"
}
| null | null | null | null | true | null |
14289
| null |
Default
| null | null |
null |
{
"abstract": " We prove a recognition principle for motivic infinite P1-loop spaces over a\nperfect field. This is achieved by developing a theory of framed motivic\nspaces, which is a motivic analogue of the theory of E-infinity-spaces. A\nframed motivic space is a motivic space equipped with transfers along finite\nsyntomic morphisms with trivialized cotangent complex in K-theory. Our main\nresult is that grouplike framed motivic spaces are equivalent to the full\nsubcategory of motivic spectra generated under colimits by suspension spectra.\nAs a consequence, we deduce some representability results for suspension\nspectra of smooth varieties, and in particular for the motivic sphere spectrum,\nin terms of Hilbert schemes of points in affine spaces.\n",
"title": "Motivic infinite loop spaces"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14290
| null |
Validated
| null | null |
null |
{
"abstract": " Correlation functions of dimer operators, the product operators of spins on\ntwo adjacent sites, are studied in the spin-$\\frac{1}{2}$ XXZ chain in the\ncritical regime. The amplitudes of the leading oscillating terms in the dimer\ncorrelation functions are determined with high accuracy as functions of the\nexchange anisotropy parameter and the external magnetic field, through the\ncombined use of bosonization and density-matrix renormalization group methods.\nIn particular, for the antiferromagnetic Heisenberg model with SU(2) symmetry,\nlogarithmic corrections to the dimer correlations due to the\nmarginally-irrelevant operator are studied, and the asymptotic form of the\ndimer correlation function is obtained. The asymptotic form of the spin-Peierls\nexcitation gap including logarithmic corrections is also derived.\n",
"title": "Dimer correlation amplitudes and dimer excitation gap in spin-1/2 XXZ and Heisenberg chains"
}
| null | null | null | null | true | null |
14291
| null |
Default
| null | null |
null |
{
"abstract": " With the advent of the 5th generation of wireless standards and an increasing\ndemand for higher throughput, methods to improve the spectral efficiency of\nwireless systems have become very important. In the context of cognitive radio,\na substantial increase in throughput is possible if the secondary user can make\nsmart decisions regarding which channel to sense and when or how often to\nsense. Here, we propose an algorithm to not only select a channel for data\ntransmission but also to predict how long the channel will remain unoccupied so\nthat the time spent on channel sensing can be minimized. Our algorithm learns\nin two stages - a reinforcement learning approach for channel selection and a\nBayesian approach to determine the optimal duration for which sensing can be\nskipped. Comparisons with other learning methods are provided through extensive\nsimulations. We show that the number of sensing is minimized with negligible\nincrease in primary interference; this implies that lesser energy is spent by\nthe secondary user in sensing and also higher throughput is achieved by saving\non sensing.\n",
"title": "Spectrum Access In Cognitive Radio Using A Two Stage Reinforcement Learning Approach"
}
| null | null | null | null | true | null |
14292
| null |
Default
| null | null |
null |
{
"abstract": " Fullerenes have attracted interest for their possible applications in various\nelectronic, biological, and optoelectronic devices. However, for efficient use\nin such devices, a suitable anchoring group has to be employed that forms\nwell-defined and stable contacts with the electrodes. In this work, we propose\na novel fullerene tetramalonate derivate functionalized with trans-1\n4,5-diazafluorene anchoring groups. The conductance of single-molecule\njunctions, investigated in two different setups with the mechanically\ncontrolled break junction technique, reveals the formation of molecular\njunctions at three conductance levels. We attribute the conductance peaks to\nthree binding modes of the anchoring groups to the gold electrodes. Density\nfunctional theory calculations confirm the existence of multiple binding\nconfigurations and calculated transmission functions are consistent with\nexperimentally determined conductance values.\n",
"title": "Charge transport through a single molecule of trans-1-bis-diazofluorene [60]fullerene"
}
| null | null | null | null | true | null |
14293
| null |
Default
| null | null |
null |
{
"abstract": " Using the existing simplified model framework, we build several dark matter\nmodels which have suppressed spin-independent scattering cross section. We show\nthat the scattering cross section can vanish due to interference effects with\nmodels obtained by simple combinations of simplified models. For weakly\ninteracting massive particle (WIMP) masses $\\gtrsim$10 GeV, collider limits are\nusually much weaker than the direct detection limits coming from LUX or\nXENON100. However, for our model combinations, LHC analyses are more\ncompetitive for some parts of the parameter space. The regions with direct\ndetection blind spots can be strongly constrained from the complementary use of\nseveral Large Hadron Collider (LHC) searches like mono-jet, jets + missing\ntransverse energy, heavy vector resonance searches, etc. We evaluate the\nstrongest limits for combinations of scalar + vector, \"squark\" + vector, and\nscalar + \"squark\" mediator, and present the LHC 14 TeV projections.\n",
"title": "Blind Spots for Direct Detection with Simplified DM Models and the LHC"
}
| null | null | null | null | true | null |
14294
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study methods to estimate drivers' posture in vehicles\nusing acceleration data of wearable sensor and conduct a field test. Recently,\nsensor technologies have been progressed. Solutions of safety management to\nanalyze vital data acquired from wearable sensor and judge work status are\nproposed. To prevent huge accidents, demands for safety management of bus and\ntaxi are high. However, acceleration of vehicles is added to wearable sensor in\nvehicles, and there is no guarantee to estimate drivers' posture accurately.\nTherefore, in this paper, we study methods to estimate driving posture using\nacceleration data acquired from T-shirt type wearable sensor hitoe, conduct\nfield tests and implement a sample application.\n",
"title": "Experiments of posture estimation on vehicles using wearable acceleration sensors"
}
| null | null | null | null | true | null |
14295
| null |
Default
| null | null |
null |
{
"abstract": " In order to make a proper reaction to the collected information from internet\nof things (IoT) devices, location information of things should be available at\nthe data center. One challenge for the massive IoT networks is to identify the\nlocation map of whole sensor nodes from partially observed distance\ninformation. In this paper, we propose a matrix completion based localization\nalgorithm to reconstruct the location map of sensors using partially observed\ndistance information. From the numerical experiments, we show that the proposed\nmethod based on the modified conjugate gradient is effective in recovering the\nEuclidean distance matrix.\n",
"title": "Matrix Completion Based Localization in the Internet of Things Network"
}
| null | null | null | null | true | null |
14296
| null |
Default
| null | null |
null |
{
"abstract": " Based on properties of n-subharmonic functions we show that a complete,\nnoncompact, properly embedded hypersurface with nonnegative Ricci curvature in\nhyperbolic space has an asymptotic boundary at infinity of at most two points.\nMoreover, the presence of two points in the asymptotic boundary is a rigidity\ncondition that forces the hypersurface to be an equidistant hypersurface about\na geodesic line in hyperbolic space. This gives an affirmative answer to the\nquestion raised by Alexander and Currier in 1990.\n",
"title": "Hypersurfaces with nonnegative Ricci curvature in hyperbolic space"
}
| null | null | null | null | true | null |
14297
| null |
Default
| null | null |
null |
{
"abstract": " We propose a simple approach which, given distributed computing resources,\ncan nearly achieve the accuracy of $k$-NN prediction, while matching (or\nimproving) the faster prediction time of $1$-NN. The approach consists of\naggregating denoised $1$-NN predictors over a small number of distributed\nsubsamples. We show, both theoretically and experimentally, that small\nsubsample sizes suffice to attain similar performance as $k$-NN, without\nsacrificing the computational efficiency of $1$-NN.\n",
"title": "Achieving the time of $1$-NN, but the accuracy of $k$-NN"
}
| null | null | null | null | true | null |
14298
| null |
Default
| null | null |
null |
{
"abstract": " There is a forgetful functor from the category of generalized effect algebras\nto the category of effect algebras. We prove that this functor is a right\nadjoint and that the corresponding left adjoint is the well-known unitization\nconstruction by Hedlíková and Pulmannová. Moreover, this adjunction is\nmonadic.\n",
"title": "A note on unitizations of generalized effect algebras"
}
| null | null | null | null | true | null |
14299
| null |
Default
| null | null |
null |
{
"abstract": " We evaluate integrals of certain polynomials over spheres and balls in real\nor complex spaces. We also promote the use of the Pochhammer symbol which gives\nthe values of our integrals in compact forms.\n",
"title": "Real and Complex Integrals on Spheres and Balls"
}
| null | null | null | null | true | null |
14300
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.