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": " The main result of this paper is a discrete Lawson correspondence between\ndiscrete CMC surfaces in R^3 and discrete minimal surfaces in S^3. This is a\ncorrespondence between two discrete isothermic surfaces. We show that this\ncorrespondence is an isometry in the following sense: it preserves the metric\ncoefficients introduced previously by Bobenko and Suris for isothermic nets.\nExactly as in the smooth case, this is a correspondence between nets with the\nsame Lax matrices, and the immersion formulas also coincide with the smooth\ncase.\n",
"title": "Discrete CMC surfaces in R^3 and discrete minimal surfaces in S^3. A discrete Lawson correspondence"
} | null | null | null | null | true | null | 20601 | null | Default | null | null |
null | {
"abstract": " Networks are fundamental models for data used in practically every\napplication domain. In most instances, several implicit or explicit choices\nabout the network definition impact the translation of underlying data to a\nnetwork representation, and the subsequent question(s) about the underlying\nsystem being represented. Users of downstream network data may not even be\naware of these choices or their impacts. We propose a task-focused network\nmodel selection methodology which addresses several key challenges. Our\napproach constructs network models from underlying data and uses minimum\ndescription length (MDL) criteria for selection. Our methodology measures\nefficiency, a general and comparable measure of the network's performance of a\nlocal (i.e. node-level) predictive task of interest. Selection on efficiency\nfavors parsimonious (e.g. sparse) models to avoid overfitting and can be\napplied across arbitrary tasks and representations. We show stability,\nsensitivity, and significance testing in our methodology.\n",
"title": "Network Model Selection Using Task-Focused Minimum Description Length"
} | null | null | null | null | true | null | 20602 | null | Default | null | null |
null | {
"abstract": " By combining analytic and geometric viewpoints on the concentration of the\ncurvature of the Milnor fibre, we prove that Lipschitz homeomorphisms preserve\nthe zones of multi-scale curvature concentration as well as the gradient canyon\nstructure of holomorphic functions of two variables. This yields the first new\nLipschitz invariants after those discovered by Henry and Parusinski in 2003.\n",
"title": "Concentration of curvature and Lipschitz invariants of holomorphic functions of two variables"
} | null | null | null | null | true | null | 20603 | null | Default | null | null |
null | {
"abstract": " Realization of a short bunch beam by manipulating the longitudinal phase\nspace distribution with a finite longitudinal dispersion following an off-crest\naccelera- tion is a widely used technique. The technique was applied in a\ncompact test accelerator of an energy-recovery linac scheme for compressing the\nbunch length at the return loop. A diagnostic system utilizing coherent\ntransition radiation was developed for the beam tuning and for estimating the\nbunch length. By scanning the beam parameters, we experimentally found the best\ncondition for the bunch compression. The RMS bunch length of 250+-50 fs was\nobtained at a bunch charge of 2 pC. This result confirmed the design and the\ntuning pro- cedure of the bunch compression operation for the future\nenergy-recovery linac (ERL).\n",
"title": "Beam tuning and bunch length measurement in the bunch compression operation at the cERL"
} | null | null | null | null | true | null | 20604 | null | Default | null | null |
null | {
"abstract": " We create fermionic dipolar $^{23}$Na$^6$Li molecules in their triplet ground\nstate from an ultracold mixture of $^{23}$Na and $^6$Li. Using\nmagneto-association across a narrow Feshbach resonance followed by a two-photon\nSTIRAP transfer to the triplet ground state, we produce $3\\,{\\times}\\,10^4$\nground state molecules in a spin-polarized state. We observe a lifetime of\n$4.6\\,\\text{s}$ in an isolated molecular sample, approaching the $p$-wave\nuniversal rate limit. Electron spin resonance spectroscopy of the triplet state\nwas used to determine the hyperfine structure of this previously unobserved\nmolecular state.\n",
"title": "Long-Lived Ultracold Molecules with Electric and Magnetic Dipole Moments"
} | null | null | null | null | true | null | 20605 | null | Default | null | null |
null | {
"abstract": " Let $\\sigma$ be arc-length measure on $S^1\\subset \\mathbb R^2$ and $\\Theta$\ndenote rotation by an angle $\\theta \\in (0, \\pi]$. Define a model bilinear\ngeneralized Radon transform, $$B_{\\theta}(f,g)(x)=\\int_{S^1} f(x-y)g(x-\\Theta\ny)\\, d\\sigma(y),$$ an analogue of the linear generalized Radon transforms of\nGuillemin and Sternberg \\cite{GS} and Phong and Stein (e.g.,\n\\cite{PhSt91,St93}). Operators such as $B_\\theta$ are motivated by problems in\ngeometric measure theory and combinatorics. For $\\theta<\\pi$, we show that\n$B_{\\theta}: L^p({\\Bbb R}^2) \\times L^q({\\Bbb R}^2) \\to L^r({\\Bbb R}^2)$ if\n$\\left(\\frac{1}{p},\\frac{1}{q},\\frac{1}{r}\\right)\\in Q$, the polyhedron with\nthe vertices $(0,0,0)$, $(\\frac{2}{3}, \\frac{2}{3}, 1)$, $(0, \\frac{2}{3},\n\\frac{1}{3})$, $(\\frac{2}{3},0,\\frac{1}{3})$, $(1,0,1)$, $(0,1,1)$ and\n$(\\frac{1}{2},\\frac{1}{2},\\frac{1}{2})$, except for $\\left(\n\\frac{1}{2},\\frac{1}{2},\\frac{1}{2} \\right)$, where we obtain a restricted\nstrong type estimate. For the degenerate case $\\theta=\\pi$, a more restrictive\nset of exponents holds. In the scale of normed spaces, $p,q,r \\ge 1$, the type\nset $Q$ is sharp. Estimates for the same exponents are also proved for a class\nof bilinear generalized Radon transforms in $\\mathbb R^2$ of the form $$\nB(f,g)(x)=\\int \\int \\delta(\\phi_1(x,y)-t_1)\\delta(\\phi_2(x,z)-t_2)\n\\delta(\\phi_3(y,z)-t_3) f(y)g(z) \\psi(y,z) \\, dy\\, dz, $$ where $\\delta$\ndenotes the Dirac distribution, $t_1,t_2,t_3\\in\\mathbb R$, $\\psi$ is a smooth\ncut-off and the defining functions $\\phi_j$ satisfy some natural geometric\nassumptions.\n",
"title": "Bilinear generalized Radon transforms in the plane"
} | null | null | null | null | true | null | 20606 | null | Default | null | null |
null | {
"abstract": " We consider the diffusion of new products in the discrete Bass-SIR model, in\nwhich consumers who adopt the product can later \"recover\" and stop influencing\ntheir peers to adopt the product. To gain insight into the effect of the social\nnetwork structure on the diffusion, we focus on two extreme cases. In the\n\"most-connected\" configuration where all consumers are inter-connected\n(complete network), averaging over all consumers leads to an aggregate model,\nwhich combines the Bass model for diffusion of new products with the SIR model\nfor epidemics. In the \"least-connected\" configuration where consumers are\narranged on a circle and each consumer can only be influenced by his left\nneighbor (one-sided 1D network), averaging over all consumers leads to a\ndifferent aggregate model which is linear, and can be solved explicitly. We\nconjecture that for any other network, the diffusion is bounded from below and\nfrom above by that on a one-sided 1D network and on a complete network,\nrespectively. When consumers are arranged on a circle and each consumer can be\ninfluenced by his left and right neighbors (two-sided 1D network), the\ndiffusion is strictly faster than on a one-sided 1D network. This is different\nfrom the case of non-recovering adopters, where the diffusion on one-sided and\non two-sided 1D networks is identical. We also propose a nonlinear model for\nrecoveries, and show that consumers' heterogeneity has a negligible effect on\nthe aggregate diffusion.\n",
"title": "Diffusion of new products with recovering consumers"
} | null | null | null | null | true | null | 20607 | null | Default | null | null |
null | {
"abstract": " While there has been substantial progress in factoid question-answering (QA),\nanswering complex questions remains challenging, typically requiring both a\nlarge body of knowledge and inference techniques. Open Information Extraction\n(Open IE) provides a way to generate semi-structured knowledge for QA, but to\ndate such knowledge has only been used to answer simple questions with\nretrieval-based methods. We overcome this limitation by presenting a method for\nreasoning with Open IE knowledge, allowing more complex questions to be\nhandled. Using a recently proposed support graph optimization framework for QA,\nwe develop a new inference model for Open IE, in particular one that can work\neffectively with multiple short facts, noise, and the relational structure of\ntuples. Our model significantly outperforms a state-of-the-art structured\nsolver on complex questions of varying difficulty, while also removing the\nreliance on manually curated knowledge.\n",
"title": "Answering Complex Questions Using Open Information Extraction"
} | null | null | null | null | true | null | 20608 | null | Default | null | null |
null | {
"abstract": " A quantum computer (QC) can solve many computational problems more\nefficiently than a classic one. The field of QCs is growing: companies (such as\nDWave, IBM, Google, and Microsoft) are building QC offerings. We position that\nsoftware engineers should look into defining a set of software engineering\npractices that apply to QC's software. To start this process, we give examples\nof challenges associated with testing such software and sketch potential\nsolutions to some of these challenges.\n",
"title": "On Testing Quantum Programs"
} | null | null | null | null | true | null | 20609 | null | Default | null | null |
null | {
"abstract": " Recurrence networks are powerful tools used effectively in the nonlinear\nanalysis of time series data. The analysis in this context is done mostly with\nunweighted and undirected complex networks constructed with specific criteria\nfrom the time series. In this work, we propose a novel method to construct\n\"weighted recurrence network\"(WRN) from a time series and show how it can\nreveal useful information regarding the structure of a chaotic attractor, which\nthe usual unweighted recurrence network cannot provide. Especially, we find the\nnode strength distribution of the WRN, from every chaotic attractor follows a\npower law (with exponential tail) with the index characteristic to the fractal\nstructure of the attractor. This leads to a new class among complex networks,\nto which networks from all standard chaotic attractors are found to belong. In\naddition, we present generalized definitions for clustering coefficient and\ncharacteristic path length and show that these measures can effectively\ndiscriminate chaotic dynamics from white noise and $1/f$ colored noise. Our\nresults indicate that the WRN and the associated measures can become\npotentially important tools for the analysis of short and noisy time series\nfrom the real world systems as they are clearly demarked from that of noisy or\nstochastic systems.\n",
"title": "Degree weighted recurrence networks for the analysis of time series data"
} | null | null | null | null | true | null | 20610 | null | Default | null | null |
null | {
"abstract": " As Ocean General Circulation Models (OGCMs) move into the petascale age,\nwhere the output from global high-resolution model runs can be of the order of\nhundreds of terabytes in size, tools to analyse the output of these models will\nneed to scale up too. Lagrangian Ocean Analysis, where virtual particles are\ntracked through hydrodynamic fields, is an increasingly popular way to analyse\nOGCM output, by mapping pathways and connectivity of biotic and abiotic\nparticulates. However, the current software stack of Lagrangian Ocean Analysis\ncodes is not dynamic enough to cope with the increasing complexity, scale and\nneed for customisation of use-cases. Furthermore, most community codes are\ndeveloped for stand-alone use, making it a nontrivial task to integrate virtual\nparticles at runtime of the OGCM. Here, we introduce the new Parcels code,\nwhich was designed from the ground up to be sufficiently scalable to cope with\npetascale computing. We highlight its API design that combines flexibility and\ncustomisation with the ability to optimise for HPC workflows, following the\nparadigm of domain-specific languages. Parcels is primarily written in Python,\nutilising the wide range of tools available in the scientific Python ecosystem,\nwhile generating low-level C-code and using Just-In-Time compilation for\nperformance-critical computation. We show a worked-out example of its API, and\nvalidate the accuracy of the code against seven idealised test cases. This\nversion~0.9 of Parcels is focussed on laying out the API, with future work\nconcentrating on optimisation, efficiency and at-runtime coupling with OGCMs.\n",
"title": "Parcels v0.9: prototyping a Lagrangian Ocean Analysis framework for the petascale age"
} | null | null | [
"Computer Science",
"Physics"
]
| null | true | null | 20611 | null | Validated | null | null |
null | {
"abstract": " While accelerators such as GPUs have limited memory, deep neural networks are\nbecoming larger and will not fit with the memory limitation of accelerators for\ntraining. We propose an approach to tackle this problem by rewriting the\ncomputational graph of a neural network, in which swap-out and swap-in\noperations are inserted to temporarily store intermediate results on CPU\nmemory. In particular, we first revise the concept of a computational graph by\ndefining a concrete semantics for variables in a graph. We then formally show\nhow to derive swap-out and swap-in operations from an existing graph and\npresent rules to optimize the graph. To realize our approach, we developed a\nmodule in TensorFlow, named TFLMS. TFLMS is published as a pull request in the\nTensorFlow repository for contributing to the TensorFlow community. With TFLMS,\nwe were able to train ResNet-50 and 3DUnet with 4.7x and 2x larger batch size,\nrespectively. In particular, we were able to train 3DUNet using images of size\nof $192^3$ for image segmentation, which, without TFLMS, had been done only by\ndividing the images to smaller images, which affects the accuracy.\n",
"title": "TFLMS: Large Model Support in TensorFlow by Graph Rewriting"
} | null | null | null | null | true | null | 20612 | null | Default | null | null |
null | {
"abstract": " Understanding the relationship between the structure of light-harvesting\nsystems and their excitation energy transfer properties is of fundamental\nimportance in many applications including the development of next generation\nphotovoltaics. Natural light harvesting in photosynthesis shows remarkable\nexcitation energy transfer properties, which suggests that pigment-protein\ncomplexes could serve as blueprints for the design of nature inspired devices.\nMechanistic insights into energy transport dynamics can be gained by leveraging\nnumerically involved propagation schemes such as the hierarchical equations of\nmotion (HEOM). Solving these equations, however, is computationally costly due\nto the adverse scaling with the number of pigments. Therefore virtual\nhigh-throughput screening, which has become a powerful tool in material\ndiscovery, is less readily applicable for the search of novel excitonic\ndevices. We propose the use of artificial neural networks to bypass the\ncomputational limitations of established techniques for exploring the\nstructure-dynamics relation in excitonic systems. Once trained, our neural\nnetworks reduce computational costs by several orders of magnitudes. Our\npredicted transfer times and transfer efficiencies exhibit similar or even\nhigher accuracies than frequently used approximate methods such as secular\nRedfield theory\n",
"title": "Machine Learning for Quantum Dynamics: Deep Learning of Excitation Energy Transfer Properties"
} | null | null | null | null | true | null | 20613 | null | Default | null | null |
null | {
"abstract": " We study stochastic multi-armed bandits with many players. The players do not\nknow the number of players, cannot communicate with each other and if multiple\nplayers select a common arm they collide and none of them receive any reward.\nWe consider the static scenario, where the number of players remains fixed, and\nthe dynamic scenario, where the players enter and leave at any time. We provide\nalgorithms based on a novel `trekking approach' that guarantees constant regret\nfor the static case and sub-linear regret for the dynamic case with high\nprobability. The trekking approach eliminates the need to estimate the number\nof players resulting in fewer collisions and improved regret performance\ncompared to the state-of-the-art algorithms. We also develop an epoch-less\nalgorithm that eliminates any requirement of time synchronization across the\nplayers provided each player can detect the presence of other players on an\narm. We validate our theoretical guarantees using simulation based and real\ntest-bed based experiments.\n",
"title": "Multi-Player Bandits: A Trekking Approach"
} | null | null | null | null | true | null | 20614 | null | Default | null | null |
null | {
"abstract": " High-resolution non-invasive 3D study of intact spine and spinal cord\nmorphology on the level of complex vascular and neuronal organization is a\ncrucial issue for the development of treatments for the injuries and\npathologies of central nervous system (CNS). X-ray phase contrast tomography\nenables high quality 3D visualization in ex-vivo mouse model of both vascular\nand neuronal network of the soft spinal cord tissue at the scale from\nmillimeters to hundreds of nanometers without any contrast agents and\nsectioning. Until now, 3D high resolution visualization of spinal cord mostly\nhas been limited by imaging of organ extracted from vertebral column because\nhigh absorbing boney tissue drastically reduces the morphological details of\nsoft tissue in image. However, the extremely destructive procedure of bones\nremoval leads to sample deterioration and, therefore, to the lack of\nconsiderable part of information about the object. In this work we present the\ndata analysis procedure to get high resolution and high contrast 3D images of\nintact mice spinal cord surrounded by vertebras, preserving all richness of\nmicro-details of the spinal cord inhabiting inside. Our results are the first\nstep forward to the difficult way toward the high- resolution investigation of\nin-vivo model central nervous system.\n",
"title": "High-resolution investigation of spinal cord and spine"
} | null | null | null | null | true | null | 20615 | null | Default | null | null |
null | {
"abstract": " There are a number of articles which deal with Bohr's phenomenon whereas only\na few papers appeared in the literature on Rogosinski's radii for analytic\nfunctions defined on the unit disk $|z|<1$. In this article, we introduce and\ninvestigate Bohr-Rogosinski's radii for analytic functions defined for $|z|<1$.\nAlso, we prove several different improved versions of the classical Bohr's\ninequality. Finally, we also discuss the Bohr-Rogosinski's radius for a class\nof subordinations. All the results are proved to be sharp.\n",
"title": "Bohr--Rogosinski radius for analytic functions"
} | null | null | [
"Mathematics"
]
| null | true | null | 20616 | null | Validated | null | null |
null | {
"abstract": " We present a class of simple algorithms that allows to find the reaction path\nin systems with a complex potential energy landscape. The approach does not\nneed any knowledge on the product state and does not require the calculation of\nany second derivatives. The underlying idea is to use two nearby points in\nconfiguration space to locate the path of slowest ascent. By introducing a weak\nnoise term, the algorithm is able to find even low-lying saddle points that are\nnot reachable by means of a slowest ascent path. Since the algorithm makes only\nuse of the value of the potential and its gradient, the computational effort to\nfind saddles is linear in the number of degrees of freedom, if the potential is\nshort-ranged. We test the performance of the algorithm for two potential energy\nlandscapes. For the Müller-Brown surface we find that the algorithm always\nfinds the correct saddle point. For the modified Müller-Brown surface, which\nhas a saddle point that is not reachable by means of a slowest ascent path, the\nalgorithm is still able to find this saddle point with high probability.\n",
"title": "Methods to locate Saddle Points in Complex Landscapes"
} | null | null | null | null | true | null | 20617 | null | Default | null | null |
null | {
"abstract": " In an increasingly polarized world, demagogues who reduce complexity down to\nsimple arguments based on emotion are gaining in popularity. Are opinions and\nonline discussions falling into demagoguery? In this work, we aim to provide\ncomputational tools to investigate this question and, by doing so, explore the\nnature and complexity of online discussions and their space of opinions,\nuncovering where each participant lies.\nMore specifically, we present a modeling framework to construct latent\nrepresentations of opinions in online discussions which are consistent with\nhuman judgements, as measured by online voting. If two opinions are close in\nthe resulting latent space of opinions, it is because humans think they are\nsimilar. Our modeling framework is theoretically grounded and establishes a\nsurprising connection between opinions and voting models and the sign-rank of a\nmatrix. Moreover, it also provides a set of practical algorithms to both\nestimate the dimension of the latent space of opinions and infer where opinions\nexpressed by the participants of an online discussion lie in this space.\nExperiments on a large dataset from Yahoo! News, Yahoo! Finance, Yahoo! Sports,\nand the Newsroom app suggest that unidimensional opinion models may often be\nunable to accurately represent online discussions, provide insights into human\njudgements and opinions, and show that our framework is able to circumvent\nlanguage nuances such as sarcasm or humor by relying on human judgements\ninstead of textual analysis.\n",
"title": "On the Complexity of Opinions and Online Discussions"
} | null | null | null | null | true | null | 20618 | null | Default | null | null |
null | {
"abstract": " We study the semi-discrete directed polymer model introduced by O'Connell-Yor\nin its stationary regime, based on our previous work on the stationary\n$q$-totally asymmetric simple exclusion process ($q$-TASEP) using a two-sided\n$q$-Whittaker process. We give a formula for the free energy distribution of\nthe polymer model in terms of Fredholm determinant and show that the universal\nKPZ stationary distribution appears in the long time limit. We also consider\nthe limit to the stationary KPZ equation and discuss the connections with\npreviously found formulas.\n",
"title": "Free energy distribution of the stationary O'Connell-Yor directed random polymer model"
} | null | null | null | null | true | null | 20619 | null | Default | null | null |
null | {
"abstract": " In this paper, we introduce the use of a personalized Gaussian Process model\n(pGP) to predict per-patient changes in ADAS-Cog13 -- a significant predictor\nof Alzheimer's Disease (AD) in the cognitive domain -- using data from each\npatient's previous visits, and testing on future (held-out) data. We start by\nlearning a population-level model using multi-modal data from previously seen\npatients using a base Gaussian Process (GP) regression. The personalized GP\n(pGP) is formed by adapting the base GP sequentially over time to a new\n(target) patient using domain adaptive GPs. We extend this personalized\napproach to predict the values of ADAS-Cog13 over the future 6, 12, 18, and 24\nmonths. We compare this approach to a GP model trained only on past data of the\ntarget patients (tGP), as well as to a new approach that combines pGP with tGP.\nWe find that the new approach, combining pGP with tGP, leads to large\nimprovements in accurately forecasting future ADAS-Cog13 scores.\n",
"title": "Personalized Gaussian Processes for Forecasting of Alzheimer's Disease Assessment Scale-Cognition Sub-Scale (ADAS-Cog13)"
} | null | null | null | null | true | null | 20620 | null | Default | null | null |
null | {
"abstract": " The paper provides results for a non-standard, hyperbolic, 1-D, nonlinear\ntraffic flow model on a bounded domain. The model consists of two first-order\nPDEs with a dynamic boundary condition that involves the time derivative of the\nvelocity. The proposed model has features that are important from a\ntraffic-theoretic point of view: is completely anisotropic and information\ntravels forward exactly at the same speed as traffic. It is shown that, for all\nphysically meaningful initial conditions, the model admits a globally defined,\nunique, classical solution that remains positive and bounded for all times.\nMoreover, it is shown that global stabilization can be achieved for arbitrary\nequilibria by means of an explicit boundary feedback law. The stabilizing\nfeedback law depends only on the inlet velocity and consequently, the\nmeasurement requirements for the implementation of the proposed boundary\nfeedback law are minimal. The efficiency of the proposed boundary feedback law\nis demonstrated by means of a numerical example.\n",
"title": "Analysis and Control of a Non-Standard Hyperbolic PDE Traffic Flow Model"
} | null | null | [
"Computer Science",
"Mathematics"
]
| null | true | null | 20621 | null | Validated | null | null |
null | {
"abstract": " Unsupervised learning techniques in computer vision often require learning\nlatent representations, such as low-dimensional linear and non-linear\nsubspaces. Noise and outliers in the data can frustrate these approaches by\nobscuring the latent spaces.\nOur main goal is deeper understanding and new development of robust\napproaches for representation learning. We provide a new interpretation for\nexisting robust approaches and present two specific contributions: a new robust\nPCA approach, which can separate foreground features from dynamic background,\nand a novel robust spectral clustering method, that can cluster facial images\nwith high accuracy. Both contributions show superior performance to standard\nmethods on real-world test sets.\n",
"title": "Learning Robust Representations for Computer Vision"
} | null | null | null | null | true | null | 20622 | null | Default | null | null |
null | {
"abstract": " In this work, we explore the problems of detecting the number of narrow-band,\nfar-field targets and estimating their corresponding directions from single\nsnapshot measurements. The principles of sparse signal recovery (SSR) are used\nfor the single snapshot detection and estimation of multiple targets. In the\nSSR framework, the DoA estimation problem is grid based and can be posed as the\nlasso optimization problem. However, the SSR framework for DoA estimation gives\nrise to the grid mismatch problem, when the unknown targets (sources) are not\nmatched with the estimation grid chosen for the construction of the array\nsteering matrix at the receiver. The block sparse recovery framework is known\nto mitigate the grid mismatch problem by jointly estimating the targets and\ntheir corresponding offsets from the estimation grid using the group lasso\nestimator. The corresponding detection problem reduces to estimating the\noptimal regularization parameter ($\\tau$) of the lasso (in case of perfect\ngrid-matching) or group-lasso estimation problem for achieving the required\nprobability of correct detection ($P_c$). We propose asymptotic and finite\nsample test statistics for detecting the number of sources with the required\n$P_c$ at moderate to high signal to noise ratios. Once the number of sources\nare detected, or equivalently the optimal $\\hat{\\tau}$ is estimated, the\ncorresponding estimation and grid matching of the DoAs can be performed by\nsolving the lasso or group-lasso problem at $\\hat{\\tau}$\n",
"title": "Detection Estimation and Grid matching of Multiple Targets with Single Snapshot Measurements"
} | null | null | null | null | true | null | 20623 | null | Default | null | null |
null | {
"abstract": " The latent feature relational model (LFRM) is a generative model for\ngraph-structured data to learn a binary vector representation for each node in\nthe graph. The binary vector denotes the node's membership in one or more\ncommunities. At its core, the LFRM miller2009nonparametric is an overlapping\nstochastic blockmodel, which defines the link probability between any pair of\nnodes as a bilinear function of their community membership vectors. Moreover,\nusing a nonparametric Bayesian prior (Indian Buffet Process) enables learning\nthe number of communities automatically from the data. However, despite its\nappealing properties, inference in LFRM remains a challenge and is typically\ndone via MCMC methods. This can be slow and may take a long time to converge.\nIn this work, we develop a small-variance asymptotics based framework for the\nnon-parametric Bayesian LFRM. This leads to an objective function that retains\nthe nonparametric Bayesian flavor of LFRM, while enabling us to design\ndeterministic inference algorithms for this model, that are easy to implement\n(using generic or specialized optimization routines) and are fast in practice.\nOur results on several benchmark datasets demonstrate that our algorithm is\ncompetitive to methods such as MCMC, while being much faster.\n",
"title": "Small-Variance Asymptotics for Nonparametric Bayesian Overlapping Stochastic Blockmodels"
} | null | null | [
"Statistics"
]
| null | true | null | 20624 | null | Validated | null | null |
null | {
"abstract": " The ancient mind/body problem continues to be one of deepest mysteries of\nscience and of the human spirit. Despite major advances in many fields, there\nis still no plausible link between subjective experience (qualia) and its\nrealization in the body. This paper outlines some of the elements of a rigorous\nscience of mind (SoM) - key ideas include scientific realism of mind, agnostic\nmysterianism, careful attention to language, and a focus on concrete\n(touchstone) questions and results.\n",
"title": "Towards a Science of Mind"
} | null | null | null | null | true | null | 20625 | null | Default | null | null |
null | {
"abstract": " We investigate the dependence of transmission losses on the choice of a slack\nbus in high voltage AC transmission networks. We formulate a transmission loss\nminimization problem in terms of slack variables representing the additional\npower injection that each generator provides to compensate the transmission\nlosses. We show analytically that for transmission lines having small,\nhomogeneous resistance over reactance ratios ${r/x\\ll1}$, transmission losses\nare generically minimal in the case of a unique \\textit{slack bus} instead of a\ndistributed slack bus. For the unique slack bus scenario, to lowest order in\n${r/x}$, transmission losses depend linearly on a resistance distance based\nindicator measuring the separation of the slack bus candidate from the rest of\nthe network. We confirm these results numerically for several IEEE and Pegase\ntestcases, and show that our predictions qualitatively hold also in the case of\nlines having inhomogeneous ${r/x}$ ratios, with optimal slack bus choices\nreducing transmission losses by ${10}\\%$ typically.\n",
"title": "Resistance distance criterion for optimal slack bus selection"
} | null | null | null | null | true | null | 20626 | null | Default | null | null |
null | {
"abstract": " Standard interpolation techniques are implicitly based on the assumption that\nthe signal lies on a homogeneous domain. In this letter, the proposed\ninterpolation method instead exploits prior information about domain\ninhomogeneity, characterized by different, potentially overlapping, subdomains.\nBy introducing a domain-similarity metric for each sample, the interpolation\nprocess is then based on a domain-informed consistency principle. We illustrate\nand demonstrate the feasibility of domain-informed linear interpolation in 1D,\nand also, on a real fMRI image in 2D. The results show the benefit of\nincorporating domain knowledge so that, for example, sharp domain boundaries\ncan be recovered by the interpolation, if such information is available.\n",
"title": "Interpolation in the Presence of Domain Inhomogeneity"
} | null | null | null | null | true | null | 20627 | null | Default | null | null |
null | {
"abstract": " This is an elementary introduction to infinite-dimensional probability. In\nthe lectures, we compute the exact mean values of some functionals on C[0,1]\nand L[0,1] by considering these functionals as infinite-dimensional random\nvariables. The results show that there exist the complete concentration of\nmeasure phenomenon for these mean values since the variances are all zeroes.\n",
"title": "Lectures on the mean values of functionals -- An elementary introduction to infinite-dimensional probability"
} | null | null | [
"Physics",
"Mathematics",
"Statistics"
]
| null | true | null | 20628 | null | Validated | null | null |
null | {
"abstract": " Repair mechanisms are important within resilient systems to maintain the\nsystem in an operational state after an error occurred. Usually, constraints on\nthe repair mechanisms are imposed, e.g., concerning the time or resources\nrequired (such as energy consumption or other kinds of costs). For systems\nmodeled by Markov decision processes (MDPs), we introduce the concept of\nresilient schedulers, which represent control strategies guaranteeing that\nthese constraints are always met within some given probability. Assigning\nrewards to the operational states of the system, we then aim towards resilient\nschedulers which maximize the long-run average reward, i.e., the expected mean\npayoff. We present a pseudo-polynomial algorithm that decides whether a\nresilient scheduler exists and if so, yields an optimal resilient scheduler. We\nshow also that already the decision problem asking whether there exists a\nresilient scheduler is PSPACE-hard.\n",
"title": "Synthesis of Optimal Resilient Control Strategies"
} | null | null | null | null | true | null | 20629 | null | Default | null | null |
null | {
"abstract": " Every real is computable from a Martin-Loef random real. This well known\nresult in algorithmic randomness was proved by Kucera and Gacs. In this survey\narticle we discuss various approaches to the problem of coding an arbitrary\nreal into a Martin-Loef random real,and also describe new results concerning\noptimal methods of coding. We start with a simple presentation of the original\nmethods of Kucera and Gacs and then rigorously demonstrate their limitations in\nterms of the size of the redundancy in the codes that they produce. Armed with\na deeper understanding of these methods, we then proceed to motivate and\nillustrate aspects of the new coding method that was recently introduced by\nBarmpalias and Lewis-Pye and which achieves optimal logarithmic redundancy, an\nexponential improvement over the original redundancy bounds.\n",
"title": "Limits of the Kucera-Gacs coding method"
} | null | null | null | null | true | null | 20630 | null | Default | null | null |
null | {
"abstract": " This paper proposes the use of subspace tracking algorithms for performing\nMIMO channel estimation at millimeter wave (mmWave) frequencies. Using a\nsubspace approach, we develop a protocol enabling the estimation of the right\n(resp. left) singular vectors at the transmitter (resp. receiver) side; then,\nwe adapt the projection approximation subspace tracking with deflation (PASTd)\nand the orthogonal Oja (OOJA) algorithms to our framework and obtain two\nchannel estimation algorithms. The hybrid analog/digital nature of the\nbeamformer is also explicitly taken into account at the algorithm design stage.\nNumerical results show that the proposed estimation algorithms are effective,\nand that they perform better than two relevant competing alternatives available\nin the open literature.\n",
"title": "Subspace Tracking Algorithms for Millimeter Wave MIMO Channel Estimation with Hybrid Beamforming"
} | null | null | [
"Computer Science"
]
| null | true | null | 20631 | null | Validated | null | null |
null | {
"abstract": " We consider a particle dressed with boundary gravitons in three-dimensional\nMinkowski space. The existence of BMS transformations implies that the\nparticle's wavefunction picks up a Berry phase when subjected to changes of\nreference frames that trace a closed path in the asymptotic symmetry group. We\nevaluate this phase and show that, for BMS superrotations, it provides a\ngravitational generalization of Thomas precession. In principle, such phases\nare observable signatures of asymptotic symmetries.\n",
"title": "Thomas Precession for Dressed Particles"
} | null | null | [
"Physics",
"Mathematics"
]
| null | true | null | 20632 | null | Validated | null | null |
null | {
"abstract": " We study the behavior of exponential random graphs in both the sparse and the\ndense regime. We show that exponential random graphs are approximate mixtures\nof graphs with independent edges whose probability matrices are critical points\nof an associated functional, thereby satisfying a certain matrix equation. In\nthe dense regime, every solution to this equation is close to a block matrix,\nconcluding that the exponential random graph behaves roughly like a mixture of\nstochastic block models. We also show existence and uniqueness of solutions to\nthis equation for several families of exponential random graphs, including the\ncase where the subgraphs are counted with positive weights and the case where\nall weights are small in absolute value. In particular, this generalizes some\nof the results in a paper by Chatterjee and Diaconis from the dense regime to\nthe sparse regime and strengthens their bounds from the cut-metric to the\none-metric.\n",
"title": "Exponential random graphs behave like mixtures of stochastic block models"
} | null | null | null | null | true | null | 20633 | null | Default | null | null |
null | {
"abstract": " Proof-carrying-code was proposed as a solution to ensure a trust relationship\nbetween two parties: a (heavyweight) analyzer and a (lightweight) checker. The\nanalyzer verifies the conformance of a given application to a specified\nproperty and generates a certificate attesting the validity of the analysis\nresult. It suffices then for the checker just to test the consistency of the\nproof instead of constructing it. We set out to study the applicability of this\ntechnique in the context of data- flow analysis. In particular, we want to know\nif there is a significant performance difference between the analyzer and the\nchecker. Therefore, we developed a tool, called DCert, implementing an\ninter-procedural context and flow-sensitive data-flow analyzer and checker for\nAndroid. Applying our tool to real-world large applications, we found out that\nchecking can be up to 8 times faster than verification. This important gain in\ntime suggests a potential for equipping applications on app stores with\ncertificates that can be checked on mobile devices which are limited in\ncomputation and storage resources. We describe our implementation and report on\nexperimental results.\n",
"title": "Certificate Enhanced Data-Flow Analysis"
} | null | null | null | null | true | null | 20634 | null | Default | null | null |
null | {
"abstract": " Until now, little was known about properties of small cells in a Poisson\nhyperplane tessellation. The few existing results were either heuristic or\napplying only to the two dimensional case and for very specific size\nfunctionals and directional distributions. This paper fills this gap by\nproviding a systematic study of small cells in a Poisson hyperplane\ntessellation of arbitrary dimension, arbitrary directional distribution\n$\\varphi$ and with respect to an arbitrary size functional $\\Sigma$. More\nprecisely, we investigate the distribution of the typical cell $Z$, conditioned\non the event $\\{\\Sigma(Z)<a\\}$, where $a\\to0$ and $\\Sigma$ is a size\nfunctional, i.e. a functional on the set of convex bodies which is continuous,\nnot identically zero, homogeneous of degree $k>0$, and increasing with respect\nto set inclusion. We focus on the number of facets and the shape of such small\ncells. We show in various general settings that small cells tend to minimize\nthe number of facets and that they have a non degenerated limit shape\ndistribution which depends on the size $\\Sigma$ and the directional\ndistribution. We also exhibit a class of directional distribution for which\ncells with small inradius do not tend to minimize the number of facets.\n",
"title": "Small cells in a Poisson hyperplane tessellation"
} | null | null | null | null | true | null | 20635 | null | Default | null | null |
null | {
"abstract": " We investigate the problem of guessing a discrete random variable $Y$ under a\nprivacy constraint dictated by another correlated discrete random variable $X$,\nwhere both guessing efficiency and privacy are assessed in terms of the\nprobability of correct guessing. We define $h(P_{XY}, \\epsilon)$ as the maximum\nprobability of correctly guessing $Y$ given an auxiliary random variable $Z$,\nwhere the maximization is taken over all $P_{Z|Y}$ ensuring that the\nprobability of correctly guessing $X$ given $Z$ does not exceed $\\epsilon$. We\nshow that the map $\\epsilon\\mapsto h(P_{XY}, \\epsilon)$ is strictly increasing,\nconcave, and piecewise linear, which allows us to derive a closed form\nexpression for $h(P_{XY}, \\epsilon)$ when $X$ and $Y$ are connected via a\nbinary-input binary-output channel. For $(X^n, Y^n)$ being pairs of independent\nand identically distributed binary random vectors, we similarly define\n$\\underline{h}_n(P_{X^nY^n}, \\epsilon)$ under the assumption that $Z^n$ is also\na binary vector. Then we obtain a closed form expression for\n$\\underline{h}_n(P_{X^nY^n}, \\epsilon)$ for sufficiently large, but nontrivial\nvalues of $\\epsilon$.\n",
"title": "Privacy-Aware Guessing Efficiency"
} | null | null | null | null | true | null | 20636 | null | Default | null | null |
null | {
"abstract": " In this paper the methods of forming a travel company customer base by means\nof social networks are observed. These methods are made to involve web-users of\nthe social networks (VK.com and Facebook) for positioning of the service of the\ntravel agency \"New Europe\" on the Internet. The methods of applying the\nmaintenance activities and interests of web-users are also used. So, the main\nmethod of information exchanging in modern network society is on-line social\nnetworks. The rapid development and improvement of such information and\ncommunication technologies is a key factor in the positioning of the travel\nagency brand in the global information space. The absence of time and space\nrestrictions and the speed of spreading of the information among an aim\naudience of social networks create all the conditions for effective\npopularization of the travel agency \"New Europe\" and its service in the\nInternet.\n",
"title": "Positioning services of a travel agency in social networks"
} | null | null | null | null | true | null | 20637 | null | Default | null | null |
null | {
"abstract": " In this paper, we consider two rainfall-runoff computer models. The first\nmodel is Matlab-Simulink model which simulates runoff from windrow compost pad\n(located at the Bioconversion Center in Athens, GA) over a period of time based\non rainfall events. The second model is Soil Water Assessment Tool (SWAT) which\nestimates surface runoff in the Middle Oconee River in Athens, GA. The input\nparameter spaces of both models are sensitive and high dimensional, the model\noutput for every input combination is a time-series of runoff, and these two\ncomputer models generate a wide spectrum of outputs including some that are far\nfrom reality. In order to improve the prediction accuracy, in this paper we\npropose to apply a history matching approach for calibrating these hydrological\nmodels, which also gives better insights for improved management of these\nsystems.\n",
"title": "Inverse Mapping for Rainfall-Runoff Models using History Matching Approach"
} | null | null | [
"Statistics"
]
| null | true | null | 20638 | null | Validated | null | null |
null | {
"abstract": " Classification of imbalanced datasets is a challenging task for standard\nalgorithms. Although many methods exist to address this problem in different\nways, generating artificial data for the minority class is a more general\napproach compared to algorithmic modifications. SMOTE algorithm and its\nvariations generate synthetic samples along a line segment that joins minority\nclass instances. In this paper we propose Geometric SMOTE (G-SMOTE) as a\ngeneralization of the SMOTE data generation mechanism. G-SMOTE generates\nsynthetic samples in a geometric region of the input space, around each\nselected minority instance. While in the basic configuration this region is a\nhyper-sphere, G-SMOTE allows its deformation to a hyper-spheroid and finally to\na line segment, emulating, in the last case, the SMOTE mechanism. The\nperformance of G-SMOTE is compared against multiple standard oversampling\nalgorithms. We present empirical results that show a significant improvement in\nthe quality of the generated data when G-SMOTE is used as an oversampling\nalgorithm.\n",
"title": "Geometric SMOTE: Effective oversampling for imbalanced learning through a geometric extension of SMOTE"
} | null | null | null | null | true | null | 20639 | null | Default | null | null |
null | {
"abstract": " We consider the problems of liveness verification and liveness synthesis for\nrecursive programs. The liveness verification problem (LVP) is to decide\nwhether a given omega-context-free language is contained in a given\nomega-regular language. The liveness synthesis problem (LSP) is to compute a\nstrategy so that a given omega-context-free game, when played along the\nstrategy, is guaranteed to derive a word in a given omega-regular language. The\nproblems are known to be EXPTIME-complete and EXPTIME-complete, respectively.\nOur contributions are new algorithms with optimal time complexity. For LVP, we\ngeneralize recent lasso-finding algorithms (also known as Ramsey-based\nalgorithms) from finite to recursive programs. For LSP, we generalize a recent\nsummary-based algorithm from finite to infinite words. Lasso finding and\nsummaries have proven to be efficient in a number of implementations for the\nfinite state and finite word setting.\n",
"title": "Liveness Verification and Synthesis: New Algorithms for Recursive Programs"
} | null | null | null | null | true | null | 20640 | null | Default | null | null |
null | {
"abstract": " Inference over tails is performed by applying only the results of extreme\nvalue theory. Whilst such theory is well defined and flexible enough in the\nunivariate case, multivariate inferential methods often require the imposition\nof arbitrary constraints not fully justifed by the underlying theory. In\ncontrast, our approach uses only the constraints imposed by theory. We build on\nprevious, theoretically justified work for marginal exceedances over a high,\nunknown threshold, by combining it with flexible, semiparametric copulae\nspecifications to investigate extreme dependence. Whilst giving probabilistic\njudgements about the extreme regime of all marginal variables, our approach\nformally uses the full dataset and allows for a variety of patterns of\ndependence, be them extremal or not. A new probabilistic criterion quantifying\nthe possibility that the data exhibits asymptotic independence is introduced\nand its robustness empirically studied. Estimation of functions of interest in\nextreme value analyses is performed via MCMC algorithms. Attention is also\ndevoted to the prediction of new extreme observations. Our approach is\nevaluated through a series of simulations, applied to real data sets and\nassessed against competing approaches. Evidence demonstrates that the bulk of\nthe data does not bias and improves the inferential process for the extremal\ndependence.\n",
"title": "A semiparametric approach for bivariate extreme exceedances"
} | null | null | null | null | true | null | 20641 | null | Default | null | null |
null | {
"abstract": " The Soil Moisture Active Passive (SMAP) mission has delivered valuable\nsensing of surface soil moisture since 2015. However, it has a short time span\nand irregular revisit schedule. Utilizing a state-of-the-art time-series deep\nlearning neural network, Long Short-Term Memory (LSTM), we created a system\nthat predicts SMAP level-3 soil moisture data with atmospheric forcing,\nmodel-simulated moisture, and static physiographic attributes as inputs. The\nsystem removes most of the bias with model simulations and improves predicted\nmoisture climatology, achieving small test root-mean-squared error (<0.035) and\nhigh correlation coefficient >0.87 for over 75\\% of Continental United States,\nincluding the forested Southeast. As the first application of LSTM in\nhydrology, we show the proposed network avoids overfitting and is robust for\nboth temporal and spatial extrapolation tests. LSTM generalizes well across\nregions with distinct climates and physiography. With high fidelity to SMAP,\nLSTM shows great potential for hindcasting, data assimilation, and weather\nforecasting.\n",
"title": "Prolongation of SMAP to Spatio-temporally Seamless Coverage of Continental US Using a Deep Learning Neural Network"
} | null | null | null | null | true | null | 20642 | null | Default | null | null |
null | {
"abstract": " It has been discovered previously that the topological order parameter could\nbe identified from the topological data of the Green function, namely the\n(generalized) TKNN invariant in general dimensions, for both non-interacting\nand interacting systems. In this note, we show that this phenomena has a clear\ngeometric derivation. This proposal could be regarded as an alternative proof\nfor the identification of the corresponding topological invariant and\ntopological order parameter.\n",
"title": "Note on Green Function Formalism and Topological Invariants"
} | null | null | null | null | true | null | 20643 | null | Default | null | null |
null | {
"abstract": " We propose a new model for unsupervised document embedding. Leading existing\napproaches either require complex inference or use recurrent neural networks\n(RNN) that are difficult to parallelize. We take a different route and develop\na convolutional neural network (CNN) embedding model. Our CNN architecture is\nfully parallelizable resulting in over 10x speedup in inference time over RNN\nmodels. Parallelizable architecture enables to train deeper models where each\nsuccessive layer has increasingly larger receptive field and models longer\nrange semantic structure within the document. We additionally propose a fully\nunsupervised learning algorithm to train this model based on stochastic forward\nprediction. Empirical results on two public benchmarks show that our approach\nproduces comparable to state-of-the-art accuracy at a fraction of computational\ncost.\n",
"title": "Unsupervised Document Embedding With CNNs"
} | null | null | null | null | true | null | 20644 | null | Default | null | null |
null | {
"abstract": " The purpose of this work is to construct a model for the functional\narchitecture of the primary visual cortex (V1), based on a structure of metric\nmeasure space induced by the underlying organization of receptive profiles\n(RPs) of visual cells. In order to account for the horizontal connectivity of\nV1 in such a context, a diffusion process compatible with the geometry of the\nspace is defined following the classical approach of K.-T. Sturm. The\nconstruction of our distance function does neither require any group\nparameterization of the family of RPs, nor involve any differential structure.\nAs such, it adapts to non-parameterized sets of RPs, possibly obtained through\nnumerical procedures; it also allows to model the lateral connectivity arising\nfrom non-differential metrics such as the one induced on a pinwheel surface by\na family of filters of vanishing scale. On the other hand, when applied to the\nclassical framework of Gabor filters, this construction yields a distance\napproximating the sub-Riemannian structure proposed as a model for V1 by G.\nCitti and A. Sarti [J Math Imaging Vis 24: 307 (2006)], thus showing itself to\nbe consistent with existing cortex models.\n",
"title": "A metric model for the functional architecture of the visual cortex"
} | null | null | null | null | true | null | 20645 | null | Default | null | null |
null | {
"abstract": " We consider the inverse problem of parameter estimation in a diffuse\ninterface model for tumour growth. The model consists of a fourth-order\nCahn--Hilliard system and contains three phenomenological parameters: the\ntumour proliferation rate, the nutrient consumption rate, and the chemotactic\nsensitivity. We study the inverse problem within the Bayesian framework and\nconstruct the likelihood and noise for two typical observation settings. One\nsetting involves an infinite-dimensional data space where we observe the full\ntumour. In the second setting we observe only the tumour volume, hence the data\nspace is finite-dimensional. We show the well-posedness of the posterior\nmeasure for both settings, building upon and improving the analytical results\nin [C. Kahle and K.F. Lam, Appl. Math. Optim. (2018)]. A numerical example\ninvolving synthetic data is presented in which the posterior measure is\nnumerically approximated by the Sequential Monte Carlo approach with tempering.\n",
"title": "Bayesian parameter identification in Cahn-Hilliard models for biological growth"
} | null | null | null | null | true | null | 20646 | null | Default | null | null |
null | {
"abstract": " Hyperuniform disordered photonic materials (HDPM) are spatially correlated\ndielectric structures with unconventional optical properties. They can be\ntransparent to long-wavelength radiation while at the same time have isotropic\nband gaps in another frequency range. This phenomenon raises fundamental\nquestions concerning photon transport through disordered media. While optical\ntransparency is robust against recurrent multiple scattering, little is known\nabout other transport regimes like diffusive multiple scattering or Anderson\nlocalization. Here we investigate band gaps, and we report Anderson\nlocalization in two-dimensional stealthy HDPM using numerical simulations of\nthe density of states and optical transport statistics. To establish a unified\nview, we propose a transport phase diagram. Our results show that, depending\nonly on the degree of correlation, a dielectric material can transition from\nlocalization behavior to a bandgap crossing an intermediate regime dominated by\ntunneling between weakly coupled states.\n",
"title": "Transport Phase Diagram and Anderson Localization in Hyperuniform Disordered Photonic Materials"
} | null | null | null | null | true | null | 20647 | null | Default | null | null |
null | {
"abstract": " Scientific explanation often requires inferring maximally predictive features\nfrom a given data set. Unfortunately, the collection of minimal maximally\npredictive features for most stochastic processes is uncountably infinite. In\nsuch cases, one compromises and instead seeks nearly maximally predictive\nfeatures. Here, we derive upper-bounds on the rates at which the number and the\ncoding cost of nearly maximally predictive features scales with desired\npredictive power. The rates are determined by the fractal dimensions of a\nprocess' mixed-state distribution. These results, in turn, show how widely-used\nfinite-order Markov models can fail as predictors and that mixed-state\npredictive features offer a substantial improvement.\n",
"title": "Nearly Maximally Predictive Features and Their Dimensions"
} | null | null | null | null | true | null | 20648 | null | Default | null | null |
null | {
"abstract": " Cross-lingual representations of words enable us to reason about word meaning\nin multilingual contexts and are a key facilitator of cross-lingual transfer\nwhen developing natural language processing models for low-resource languages.\nIn this survey, we provide a comprehensive typology of cross-lingual word\nembedding models. We compare their data requirements and objective functions.\nThe recurring theme of the survey is that many of the models presented in the\nliterature optimize for the same objectives, and that seemingly different\nmodels are often equivalent modulo optimization strategies, hyper-parameters,\nand such. We also discuss the different ways cross-lingual word embeddings are\nevaluated, as well as future challenges and research horizons.\n",
"title": "A Survey Of Cross-lingual Word Embedding Models"
} | null | null | null | null | true | null | 20649 | null | Default | null | null |
null | {
"abstract": " A lattice is the integer span of some linearly independent vectors. Lattice\nproblems have many significant applications in coding theory and cryptographic\nsystems for their conjectured hardness. The Shortest Vector Problem (SVP),\nwhich is to find the shortest non-zero vector in a lattice, is one of the\nwell-known problems that are believed to be hard to solve, even with a quantum\ncomputer. In this paper we propose space-efficient classical and quantum\nalgorithms for solving SVP. Currently the best time-efficient algorithm for\nsolving SVP takes $2^{n+o(n)}$ time and $2^{n+o(n)}$ space. Our classical\nalgorithm takes $2^{2.05n+o(n)}$ time to solve SVP with only $2^{0.5n+o(n)}$\nspace. We then modify our classical algorithm to a quantum version, which can\nsolve SVP in time $2^{1.2553n+o(n)}$ with $2^{0.5n+o(n)}$ classical space and\nonly poly(n) qubits.\n",
"title": "Space-efficient classical and quantum algorithms for the shortest vector problem"
} | null | null | [
"Computer Science"
]
| null | true | null | 20650 | null | Validated | null | null |
null | {
"abstract": " This research was to design a 2.4 GHz class E Power Amplifier (PA) for health\ncare, with 0.18um Semiconductor Manufacturing International Corporation CMOS\ntechnology by using Cadence software. And also RF switch was designed at\ncadence software with power Jazz 180nm SOI process. The ultimate goal for such\napplication is to reach high performance and low cost, and between high\nperformance and low power consumption design. This paper introduces the design\nof a 2.4GHz class E power amplifier and RF switch design. PA consists of\ncascade stage with negative capacitance. This power amplifier can transmit\n16dBm output power to a 50{\\Omega} load. The performance of the power amplifier\nand switch meet the specification requirements of the desired.\n",
"title": "Low Power SI Class E Power Amplifier and RF Switch For Health Care"
} | null | null | null | null | true | null | 20651 | null | Default | null | null |
null | {
"abstract": " The study of Dense-$3$-Subhypergraph problem was initiated in Chlamt{á}c\net al. [Approx'16]. The input is a universe $U$ and collection ${\\cal S}$ of\nsubsets of $U$, each of size $3$, and a number $k$. The goal is to choose a set\n$W$ of $k$ elements from the universe, and maximize the number of sets, $S\\in\n{\\cal S}$ so that $S\\subseteq W$. The members in $U$ are called {\\em vertices}\nand the sets of ${\\cal S}$ are called the {\\em hyperedges}. This is the\nsimplest extension into hyperedges of the case of sets of size $2$ which is the\nwell known Dense $k$-subgraph problem.\nThe best known ratio for the Dense-$3$-Subhypergraph is $O(n^{0.69783..})$ by\nChlamt{á}c et al. We improve this ratio to $n^{0.61802..}$. More\nimportantly, we give a new algorithm that approximates Dense-$3$-Subhypergraph\nwithin a ratio of $\\tilde O(n/k)$, which improves the ratio of $O(n^2/k^2)$ of\nChlamt{á}c et al.\nWe prove that under the {\\em log density conjecture} (see Bhaskara et al.\n[STOC'10]) the ratio cannot be better than $\\Omega(\\sqrt{n})$ and demonstrate\nsome cases in which this optimum can be attained.\n",
"title": "Improved approximation algorithm for the Dense-3-Subhypergraph Problem"
} | null | null | null | null | true | null | 20652 | null | Default | null | null |
null | {
"abstract": " A literature survey on ontologies concerning the Security Assessment domain\nhas been carried out to uncover initiatives that aim at formalizing concepts\nfrom the Security Assessment field of research. A preliminary analysis and a\ndiscussion on the selected works are presented. Our main contribution is an\nupdated literature review, describing key characteristics, results, research\nissues, and application domains of the papers. We have also detected gaps in\nthe Security Assessment literature that could be the subject of further studies\nin the field. This work is meant to be useful for security researchers who wish\nto adopt a formal approach in their methods.\n",
"title": "A Survey of Security Assessment Ontologies"
} | null | null | null | null | true | null | 20653 | null | Default | null | null |
null | {
"abstract": " Logic-based event recognition systems infer occurrences of events in time\nusing a set of event definitions in the form of first-order rules. The Event\nCalculus is a temporal logic that has been used as a basis in event recognition\napplications, providing among others, direct connections to machine learning,\nvia Inductive Logic Programming (ILP). OLED is a recently proposed ILP system\nthat learns event definitions in the form of Event Calculus theories, in a\nsingle pass over a data stream. In this work we present a version of OLED that\nallows for distributed, online learning. We evaluate our approach on a\nbenchmark activity recognition dataset and show that we can significantly\nreduce training times, exchanging minimal information between processing nodes.\n",
"title": "Distributed Online Learning of Event Definitions"
} | null | null | null | null | true | null | 20654 | null | Default | null | null |
null | {
"abstract": " Graph isomorphism is an important computer science problem. The problem for\nthe general case is unknown to be in polynomial time. The base algorithm for\nthe general case works in quasi-polynomial time. The solutions in polynomial\ntime for some special type of classes are known. In this work, we have worked\nwith a special type of graphs. We have proposed a method to represent these\ngraphs and finding isomorphism between these graphs. The method uses a modified\nversion of the degree list of a graph and neighbourhood degree list. These\nspecial type of graphs have a property that neighbourhood degree list of any\ntwo immediate neighbours is different for every vertex.The representation\nbecomes invariant to the order in which the node was selected for giving the\nrepresentation making the isomorphism problem trivial for this case. The\nalgorithm works in $O(n^4)$ time, where n is the number of vertices present in\nthe graph. The proposed algorithm runs faster than quasi-polynomial time for\nthe graphs used in the study.\n",
"title": "Solving Graph Isomorphism Problem for a Special case"
} | null | null | null | null | true | null | 20655 | null | Default | null | null |
null | {
"abstract": " Effect modification occurs when the effect of the treatment on an outcome\nvaries according to the level of other covariates and often has important\nimplications in decision making. When there are tens or hundreds of covariates,\nit becomes necessary to use the observed data to select a simpler model for\neffect modification and then make valid statistical inference. We propose a two\nstage procedure to solve this problem. First, we use Robinson's transformation\nto decouple the nuisance parameters from the treatment effect of interest and\nuse machine learning algorithms to estimate the nuisance parameters. Next,\nafter plugging in the estimates of the nuisance parameters, we use the Lasso to\nchoose a low-complexity model for effect modification. Compared to a full model\nconsisting of all the covariates, the selected model is much more\ninterpretable. Compared to the univariate subgroup analyses, the selected model\ngreatly reduces the number of false discoveries. We show that the conditional\nselective inference for the selected model is asymptotically valid given the\nrate assumptions in classical semiparametric regression. Extensive simulation\nstudies are conducted to verify the asymptotic results and an epidemiological\napplication is used to demonstrate the method.\n",
"title": "Selective inference for effect modification via the lasso"
} | null | null | null | null | true | null | 20656 | null | Default | null | null |
null | {
"abstract": " A growing body of research focuses on computationally detecting controversial\ntopics and understanding the stances people hold on them. Yet gaps remain in\nour theoretical and practical understanding of how to define controversy, how\nit manifests, and how to measure it. In this paper, we introduce a novel\nmeasure we call \"contention\", defined with respect to a topic and a population.\nWe model contention from a mathematical standpoint. We validate our model by\nexamining a diverse set of sources: real-world polling data sets, actual voter\ndata, and Twitter coverage on several topics. In our publicly-released Twitter\ndata set of nearly 100M tweets, we examine several topics such as Brexit, the\n2016 U.S. Elections, and \"The Dress\", and cross-reference them with other\nsources. We demonstrate that the contention measure holds explanatory power for\na wide variety of observed phenomena, such as controversies over climate change\nand other topics that are well within scientific consensus. Finally, we\nre-examine the notion of controversy, and present a theoretical framework that\ndefines it in terms of population. We present preliminary evidence suggesting\nthat contention is one dimension of controversy, along with others, such as\n\"importance\". Our new contention measure, along with the hypothesized model of\ncontroversy, suggest several avenues for future work in this emerging\ninterdisciplinary research area.\n",
"title": "Is Climate Change Controversial? Modeling Controversy as Contention Within Populations"
} | null | null | null | null | true | null | 20657 | null | Default | null | null |
null | {
"abstract": " Recent work has shown that state-of-the-art models are highly vulnerable to\nadversarial perturbations of the input. We propose cowboy, an approach to\ndetecting and defending against adversarial attacks by using both the\ndiscriminator and generator of a GAN trained on the same dataset. We show that\nthe discriminator consistently scores the adversarial samples lower than the\nreal samples across multiple attacks and datasets. We provide empirical\nevidence that adversarial samples lie outside of the data manifold learned by\nthe GAN. Based on this, we propose a cleaning method which uses both the\ndiscriminator and generator of the GAN to project the samples back onto the\ndata manifold. This cleaning procedure is independent of the classifier and\ntype of attack and thus can be deployed in existing systems.\n",
"title": "Defending Against Adversarial Attacks by Leveraging an Entire GAN"
} | null | null | null | null | true | null | 20658 | null | Default | null | null |
null | {
"abstract": " Exploiting the wealth of imaging and non-imaging information for disease\nprediction tasks requires models capable of representing, at the same time,\nindividual features as well as data associations between subjects from\npotentially large populations. Graphs provide a natural framework for such\ntasks, yet previous graph-based approaches focus on pairwise similarities\nwithout modelling the subjects' individual characteristics and features. On the\nother hand, relying solely on subject-specific imaging feature vectors fails to\nmodel the interaction and similarity between subjects, which can reduce\nperformance. In this paper, we introduce the novel concept of Graph\nConvolutional Networks (GCN) for brain analysis in populations, combining\nimaging and non-imaging data. We represent populations as a sparse graph where\nits vertices are associated with image-based feature vectors and the edges\nencode phenotypic information. This structure was used to train a GCN model on\npartially labelled graphs, aiming to infer the classes of unlabelled nodes from\nthe node features and pairwise associations between subjects. We demonstrate\nthe potential of the method on the challenging ADNI and ABIDE databases, as a\nproof of concept of the benefit from integrating contextual information in\nclassification tasks. This has a clear impact on the quality of the\npredictions, leading to 69.5% accuracy for ABIDE (outperforming the current\nstate of the art of 66.8%) and 77% for ADNI for prediction of MCI conversion,\nsignificantly outperforming standard linear classifiers where only individual\nfeatures are considered.\n",
"title": "Spectral Graph Convolutions for Population-based Disease Prediction"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 20659 | null | Validated | null | null |
null | {
"abstract": " An independence system (with respect to the unknotting number) is defined for\na classical knot diagram. It is proved that the independence system is a knot\ninvariant for alternating knots. The exchange property for minimal unknotting\nsets are also discussed. It is shown that there exists an infinite family of\nknot diagrams whose corresponding independence systems are matroids. In\ncontrast, infinite families of knot diagrams exist whose independence systems\nare not matroids.\n",
"title": "An independence system as knot invariant"
} | null | null | null | null | true | null | 20660 | null | Default | null | null |
null | {
"abstract": " We consider theoretically ultracold interacting bosonic atoms confined to\nquasi-one-dimensional ladder structures formed by optical lattices and coupled\nto the field of an optical cavity. The atoms can collect a spatial phase\nimprint during a cavity-assisted tunneling along a rung via Raman transitions\nemploying a cavity mode and a transverse running wave pump beam. By adiabatic\nelimination of the cavity field we obtain an effective Hamiltonian for the\nbosonic atoms, with a self-consistency condition. Using the numerical density\nmatrix renormalization group method, we obtain a rich steady state diagram of\nself-organized steady states. Transitions between superfluid to Mott-insulating\nstates occur, on top of which we can have Meissner, vortex liquid, and vortex\nlattice phases. Also a state that explicitly breaks the symmetry between the\ntwo legs of the ladder, namely the biased-ladder phase is dynamically\nstabilized.\n",
"title": "A cavity-induced artificial gauge field in a Bose-Hubbard ladder"
} | null | null | null | null | true | null | 20661 | null | Default | null | null |
null | {
"abstract": " An overview of research on laser-plasma based acceleration of ions is given.\nThe experimental state of the art is summarized and recent progress is\ndiscussed. The basic acceleration processes are briefly reviewed with an\noutlook on hybrid mechanisms and novel concepts. Finally, we put focus on the\ndevelopment of engineered targets for enhanced acceleration and of all-optical\nmethods for beam post-acceleration and control.\n",
"title": "A Review of Laser-Plasma Ion Acceleration"
} | null | null | null | null | true | null | 20662 | null | Default | null | null |
null | {
"abstract": " In this paper, we consider estimators for an additive functional of $\\phi$,\nwhich is defined as $\\theta(P;\\phi)=\\sum_{i=1}^k\\phi(p_i)$, from $n$ i.i.d.\nrandom samples drawn from a discrete distribution $P=(p_1,...,p_k)$ with\nalphabet size $k$. We propose a minimax optimal estimator for the estimation\nproblem of the additive functional. We reveal that the minimax optimal rate is\ncharacterized by the divergence speed of the fourth derivative of $\\phi$ if the\ndivergence speed is high. As a result, we show there is no consistent estimator\nif the divergence speed of the fourth derivative of $\\phi$ is larger than\n$p^{-4}$. Furthermore, if the divergence speed of the fourth derivative of\n$\\phi$ is $p^{4-\\alpha}$ for $\\alpha \\in (0,1)$, the minimax optimal rate is\nobtained within a universal multiplicative constant as $\\frac{k^2}{(n\\ln\nn)^{2\\alpha}} + \\frac{k^{2-2\\alpha}}{n}$.\n",
"title": "Minimax Optimal Estimators for Additive Scalar Functionals of Discrete Distributions"
} | null | null | null | null | true | null | 20663 | null | Default | null | null |
null | {
"abstract": " The van der Waals heterostructures of allotropes of phosphorene (${\\alpha}$-\nand $\\beta-P$) with MoSe2 (H-, T-, ZT- and SO-MoSe2) are investigated in the\nframework of state-of-the-art density functional theory. The semiconducting\nheterostructures, $\\beta$-P /H-MoSe2 and ${\\alpha}$-P / H-MoSe2, forms\nanti-type structures with type I and type II band alignments, respectively,\nwhose bands are tunable with external electric field. ${\\alpha}$-P / ZT-MoSe2\nand ${\\alpha}$-P / SO-MoSe2 form ohmic semiconductor-metal contacts while\nSchottky barrier in $\\beta$-P / T-MoSe2 can be reduced to zero by external\nelectric field to form ohmic contact which is useful to realize\nhigh-performance devices. Simulated STM images of given heterostructures reveal\nthat ${\\alpha}$-P can be used as a capping layer to differentiate between\nvarious allotropes of underlying MoSe2. The dielectric response of considered\nheterostructures is highly anisotropic in terms of lateral and vertical\npolarization. The tunable electronic and dielectric response of van der Waals\nphosphorene/MoSe2 heterostructure may find potentials applications in the\nfabrication of optoelectronic devices.\n",
"title": "Van der Waals Heterostructures Based on Allotropes of Phosphorene and MoSe2"
} | null | null | null | null | true | null | 20664 | null | Default | null | null |
null | {
"abstract": " We provide a surprising answer to a question raised in S. Ahmad and A.C.\nLazer [2], and extend the results of that paper.\n",
"title": "On separated solutions of logistic population equation with harvesting"
} | null | null | null | null | true | null | 20665 | null | Default | null | null |
null | {
"abstract": " Scattering of obliquely incident electromagnetic waves from periodically\nspace-time modulated slabs is investigated. It is shown that such structures\noperate as nonreciprocal harmonic generators and spatial-frequency filters. For\noblique incidences, low-frequency harmonics are filtered out in the form of\nsurface waves, while high-frequency harmonics are transmitted as space waves.\nIn the quasisonic regime, where the velocity of the space-time modulation is\nclose to the velocity of the electromagnetic waves in the background medium,\nthe incident wave is strongly coupled to space-time harmonics in the forward\ndirection, while in the backward direction it exhibits low coupling to other\nharmonics. This nonreciprocity is leveraged for the realization of an\nelectromagnetic isolator in the quasisonic regime and is experimentally\ndemonstrated at microwave frequencies.\n",
"title": "Nonreciprocal Electromagnetic Scattering from a Periodically Space-Time Modulated Slab and Application to a Quasisonic Isolator"
} | null | null | null | null | true | null | 20666 | null | Default | null | null |
null | {
"abstract": " This paper describes two supervised baseline systems for the Story Cloze Test\nShared Task (Mostafazadeh et al., 2016a). We first build a classifier using\nfeatures based on word embeddings and semantic similarity computation. We\nfurther implement a neural LSTM system with different encoding strategies that\ntry to model the relation between the story and the provided endings. Our\nexperiments show that a model using representation features based on average\nword embedding vectors over the given story words and the candidate ending\nsentences words, joint with similarity features between the story and candidate\nending representations performed better than the neural models. Our best model\nachieves an accuracy of 72.42, ranking 3rd in the official evaluation.\n",
"title": "Story Cloze Ending Selection Baselines and Data Examination"
} | null | null | [
"Computer Science"
]
| null | true | null | 20667 | null | Validated | null | null |
null | {
"abstract": " This paper is concerned with a linear quadratic (LQ, for short) optimal\ncontrol problem with fixed terminal states and integral quadratic constraints.\nA Riccati equation with infinite terminal value is introduced, which is\nuniquely solvable and whose solution can be approximated by the solution for a\nsuitable unconstrained LQ problem with penalized terminal state. Using results\nfrom duality theory, the optimal control is explicitly derived by solving the\nRiccati equation together with an optimal parameter selection problem. It turns\nout that the optimal control is not only a feedback of the current state, but\nalso a feedback of the target (terminal state). Some examples are presented to\nillustrate the theory developed.\n",
"title": "Linear Quadratic Optimal Control Problems with Fixed Terminal States and Integral Quadratic Constraints"
} | null | null | null | null | true | null | 20668 | null | Default | null | null |
null | {
"abstract": " Let $X=\\{x_i:i\\in\\mathbb{Z}\\}$, $\\dots<x_{i-1}<x_i<x_{i+1}<\\dots$, be a\nsampling set which is separated by a constant $\\gamma>0$. Under certain\nconditions on $\\phi$, it is proved that if there exists a positive integer\n$\\nu$ such that\n$$\\delta_\\nu:=\\sup\\limits_{i\\in\\mathbb{Z}}(x_{i+\\nu}-x_i)<\\dfrac{\\nu}{2\\pi}\\left(\\dfrac{c_{k}^2}{M_{2k}}\\right)^{\\frac{1}{4k}},$$\nthen every function belonging to a shift-invariant space $V(\\phi)$ can be\nreconstructed stably from its nonuniform sample values\n$\\{f^{(j)}(x_i):j=0,1,\\dots, k-1, i\\in\\mathbb{Z}\\}$, where $c_k$ is a\nWirtinger-Sobolev constant and $M_{2k}$ is a constant in Bernstein-type\ninequality of $V(\\phi)$. Further, when $k=1$, the maximum gap $\\delta_\\nu<\\nu$\nis sharp for certain shift-invariant spaces.\n",
"title": "A new sampling density condition for shift-invariant spaces"
} | null | null | null | null | true | null | 20669 | null | Default | null | null |
null | {
"abstract": " The mean objective cost of uncertainty (MOCU) quantifies the performance cost\nof using an operator that is optimal across an uncertainty class of systems as\nopposed to using an operator that is optimal for a particular system.\nMOCU-based experimental design selects an experiment to maximally reduce MOCU,\nthereby gaining the greatest reduction of uncertainty impacting the operational\nobjective. The original formulation applied to finding optimal system\noperators, where optimality is with respect to a cost function, such as\nmean-square error; and the prior distribution governing the uncertainty class\nrelates directly to the underlying physical system. Here we provide a\ngeneralized MOCU and the corresponding experimental design. We then demonstrate\nhow this new formulation includes as special cases MOCU-based experimental\ndesign methods developed for materials science and genomic networks when there\nis experimental error. Most importantly, we show that the classical Knowledge\nGradient and Efficient Global Optimization experimental design procedures are\nactually implementations of MOCU-based experimental design under their modeling\nassumptions.\n",
"title": "Experimental Design via Generalized Mean Objective Cost of Uncertainty"
} | null | null | null | null | true | null | 20670 | null | Default | null | null |
null | {
"abstract": " Artificial neural networks are a popular and effective machine learning\ntechnique. Great progress has been made parallelizing the expensive training\nphase of an individual network, leading to highly specialized pieces of\nhardware, many based on GPU-type architectures, and more concurrent algorithms\nsuch as synthetic gradients. However, the training phase continues to be a\nbottleneck, where the training data must be processed serially over thousands\nof individual training runs. This work considers a multigrid reduction in time\n(MGRIT) algorithm that is able to parallelize over the thousands of training\nruns and converge to the exact same solution as traditional training would\nprovide. MGRIT was originally developed to provide parallelism for time\nevolution problems that serially step through a finite number of time-steps.\nThis work recasts the training of a neural network similarly, treating neural\nnetwork training as an evolution equation that evolves the network weights from\none step to the next. Thus, this work concerns distributed computing approaches\nfor neural networks, but is distinct from other approaches which seek to\nparallelize only over individual training runs. The work concludes with\nsupporting numerical results for two model problems.\n",
"title": "Parallelizing Over Artificial Neural Network Training Runs with Multigrid"
} | null | null | null | null | true | null | 20671 | null | Default | null | null |
null | {
"abstract": " Curating labeled training data has become the primary bottleneck in machine\nlearning. Recent frameworks address this bottleneck with generative models to\nsynthesize labels at scale from weak supervision sources. The generative\nmodel's dependency structure directly affects the quality of the estimated\nlabels, but selecting a structure automatically without any labeled data is a\ndistinct challenge. We propose a structure estimation method that maximizes the\n$\\ell_1$-regularized marginal pseudolikelihood of the observed data. Our\nanalysis shows that the amount of unlabeled data required to identify the true\nstructure scales sublinearly in the number of possible dependencies for a broad\nclass of models. Simulations show that our method is 100$\\times$ faster than a\nmaximum likelihood approach and selects $1/4$ as many extraneous dependencies.\nWe also show that our method provides an average of 1.5 F1 points of\nimprovement over existing, user-developed information extraction applications\non real-world data such as PubMed journal abstracts.\n",
"title": "Learning the Structure of Generative Models without Labeled Data"
} | null | null | null | null | true | null | 20672 | null | Default | null | null |
null | {
"abstract": " In many applications of classifier learning, training data suffers from label\nnoise. Deep networks are learned using huge training data where the problem of\nnoisy labels is particularly relevant. The current techniques proposed for\nlearning deep networks under label noise focus on modifying the network\narchitecture and on algorithms for estimating true labels from noisy labels. An\nalternate approach would be to look for loss functions that are inherently\nnoise-tolerant. For binary classification there exist theoretical results on\nloss functions that are robust to label noise. In this paper, we provide some\nsufficient conditions on a loss function so that risk minimization under that\nloss function would be inherently tolerant to label noise for multiclass\nclassification problems. These results generalize the existing results on\nnoise-tolerant loss functions for binary classification. We study some of the\nwidely used loss functions in deep networks and show that the loss function\nbased on mean absolute value of error is inherently robust to label noise. Thus\nstandard back propagation is enough to learn the true classifier even under\nlabel noise. Through experiments, we illustrate the robustness of risk\nminimization with such loss functions for learning neural networks.\n",
"title": "Robust Loss Functions under Label Noise for Deep Neural Networks"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 20673 | null | Validated | null | null |
null | {
"abstract": " Today, we have to deal with many data (Big data) and we need to make\ndecisions by choosing an architectural framework to analyze these data coming\nfrom different area. Due to this, it become problematic when we want to process\nthese data, and even more, when it is continuous data. When you want to process\nsome data, you have to first receive it, store it, and then query it. This is\nwhat we call Batch Processing. It works well when you process big amount of\ndata, but it finds its limits when you want to get fast (or real-time)\nprocessing results, such as financial trades, sensors, user session activity,\netc. The solution to this problem is stream processing. Stream processing\napproach consists of data arriving record by record and rather than storing it,\nthe processing should be done directly. Therefore, direct results are needed\nwith a latency that may vary in real-time.\nIn this paper, we propose an assessment quality model to evaluate and choose\nstream processing frameworks. We describe briefly different architectural\nframeworks such as Kafka, Spark Streaming and Flink that address the stream\nprocessing. Using our quality model, we present a decision tree to support\nengineers to choose a framework following the quality aspects. Finally, we\nevaluate our model doing a case study to Twitter and Netflix streaming.\n",
"title": "A quality model for evaluating and choosing a stream processing framework architecture"
} | null | null | null | null | true | null | 20674 | null | Default | null | null |
null | {
"abstract": " We consider a repeated newsvendor problem where the inventory manager has no\nprior information about the demand, and can access only censored/sales data. In\nanalogy to multi-armed bandit problems, the manager needs to simultaneously\n\"explore\" and \"exploit\" with her inventory decisions, in order to minimize the\ncumulative cost. We make no probabilistic assumptions---importantly,\nindependence or time stationarity---regarding the mechanism that creates the\ndemand sequence. Our goal is to shed light on the hardness of the problem, and\nto develop policies that perform well with respect to the regret criterion,\nthat is, the difference between the cumulative cost of a policy and that of the\nbest fixed action/static inventory decision in hindsight, uniformly over all\nfeasible demand sequences. We show that a simple randomized policy, termed the\nExponentially Weighted Forecaster, combined with a carefully designed cost\nestimator, achieves optimal scaling of the expected regret (up to logarithmic\nfactors) with respect to all three key primitives: the number of time periods,\nthe number of inventory decisions available, and the demand support. Through\nthis result, we derive an important insight: the benefit from \"information\nstalking\" as well as the cost of censoring are both negligible in this dynamic\nlearning problem, at least with respect to the regret criterion. Furthermore,\nwe modify the proposed policy in order to perform well in terms of the tracking\nregret, that is, using as benchmark the best sequence of inventory decisions\nthat switches a limited number of times. Numerical experiments suggest that the\nproposed approach outperforms existing ones (that are tailored to, or\nfacilitated by, time stationarity) on nonstationary demand models. Finally, we\nextend the proposed approach and its analysis to a \"combinatorial\" version of\nthe repeated newsvendor problem.\n",
"title": "On the Hardness of Inventory Management with Censored Demand Data"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 20675 | null | Validated | null | null |
null | {
"abstract": " Social networking sites such as Twitter have provided a great opportunity for\norganizations such as public libraries to disseminate information for public\nrelations purposes. However, there is a need to analyze vast amounts of social\nmedia data. This study presents a computational approach to explore the content\nof tweets posted by nine public libraries in the northeastern United States of\nAmerica. In December 2017, this study extracted more than 19,000 tweets from\nthe Twitter accounts of seven state libraries and two urban public libraries.\nComputational methods were applied to collect the tweets and discover\nmeaningful themes. This paper shows how the libraries have used Twitter to\nrepresent their services and provides a starting point for different\norganizations to evaluate the themes of their public tweets.\n",
"title": "Social Media Analysis For Organizations: Us Northeastern Public And State Libraries Case Study"
} | null | null | null | null | true | null | 20676 | null | Default | null | null |
null | {
"abstract": " We study the class of rings $R$ with the property that for $x\\in R$ at least\none of the elements $x$ and $1+x$ are tripotent.\n",
"title": "Weakly tripotent rings"
} | null | null | null | null | true | null | 20677 | null | Default | null | null |
null | {
"abstract": " Effective riverine flood forecasting at scale is hindered by a multitude of\nfactors, most notably the need to rely on human calibration in current\nmethodology, the limited amount of data for a specific location, and the\ncomputational difficulty of building continent/global level models that are\nsufficiently accurate. Machine learning (ML) is primed to be useful in this\nscenario: learned models often surpass human experts in complex\nhigh-dimensional scenarios, and the framework of transfer or multitask learning\nis an appealing solution for leveraging local signals to achieve improved\nglobal performance. We propose to build on these strengths and develop ML\nsystems for timely and accurate riverine flood prediction.\n",
"title": "ML for Flood Forecasting at Scale"
} | null | null | null | null | true | null | 20678 | null | Default | null | null |
null | {
"abstract": " We consider the frequency domain form of proper orthogonal decomposition\n(POD) called spectral proper orthogonal decomposition (SPOD). Spectral POD is\nderived from a space-time POD problem for statistically stationary flows and\nleads to modes that each oscillate at a single frequency. This form of POD goes\nback to the original work of Lumley (Stochastic tools in turbulence, Academic\nPress, 1970), but has been overshadowed by a space-only form of POD since the\n1990s. We clarify the relationship between these two forms of POD and show that\nSPOD modes represent structures that evolve coherently in space and time while\nspace-only POD modes in general do not. We also establish a relationship\nbetween SPOD and dynamic mode decomposition (DMD); we show that SPOD modes are\nin fact optimally averaged DMD modes obtained from an ensemble DMD problem for\nstationary flows. Accordingly, SPOD modes represent structures that are dynamic\nin the same sense as DMD modes but also optimally account for the statistical\nvariability of turbulent flows. Finally, we establish a connection between SPOD\nand resolvent analysis. The key observation is that the resolvent-mode\nexpansion coefficients must be regarded as statistical quantities to ensure\nconvergent approximations of the flow statistics. When the expansion\ncoefficients are uncorrelated, we show that SPOD and resolvent modes are\nidentical. Our theoretical results and the overall utility of SPOD are\ndemonstrated using two example problems: the complex Ginzburg-Landau equation\nand a turbulent jet.\n",
"title": "Spectral proper orthogonal decomposition and its relationship to dynamic mode decomposition and resolvent analysis"
} | null | null | null | null | true | null | 20679 | null | Default | null | null |
null | {
"abstract": " When stochastic dominance $F\\leq_{st}G$ does not hold, we can improve\nagreement to stochastic order by suitably trimming both distributions. In this\nwork we consider the $L_2-$Wasserstein distance, $\\mathcal W_2$, to stochastic\norder of these trimmed versions. Our characterization for that distance\nnaturally leads to consider a $\\mathcal W_2$-based index of disagreement with\nstochastic order, $\\varepsilon_{\\mathcal W_2}(F,G)$. We provide asymptotic\nresults allowing to test $H_0: \\varepsilon_{\\mathcal W_2}(F,G)\\geq\n\\varepsilon_0$ vs $H_a: \\varepsilon_{\\mathcal W_2}(F,G)<\\varepsilon_0$, that,\nunder rejection, would give statistical guarantee of almost stochastic\ndominance. We include a simulation study showing a good performance of the\nindex under the normal model.\n",
"title": "An optimal transportation approach for assessing almost stochastic order"
} | null | null | null | null | true | null | 20680 | null | Default | null | null |
null | {
"abstract": " An electron lens can serve as an effective mechanism for suppressing coherent\ninstabilities in high intensity storage rings through nonlinear amplitude\ndependent betatron tune shift. However, the addition of a strong localized\nnonlinear focusing element to the accelerator lattice may lead to undesired\neffects in particle dynamics. We evaluate the effect of a Gaussian electron\nlens on single particle motion in HL-LHC using numerical tracking simulations,\nand compare the results to the case when an equal tune spread is generated by\nconventional octupole magnets.\n",
"title": "The Effect of Electron Lens as Landau Damping Device on Single Particle Dynamics in HL-LHC"
} | null | null | [
"Physics"
]
| null | true | null | 20681 | null | Validated | null | null |
null | {
"abstract": " We search for sterile neutrinos in the holographic dark energy cosmology by\nusing the latest observational data. To perform the analysis, we employ the\ncurrent cosmological observations, including the cosmic microwave background\ntemperature power spectrum data from the Planck mission, the baryon acoustic\noscillation measurements, the type Ia supernova data, the redshift space\ndistortion measurements, the shear data of weak lensing observation, the Planck\nlensing measurement, and the latest direct measurement of $H_0$ as well. We\nshow that, compared to the $\\Lambda$CDM cosmology, the holographic dark energy\ncosmology with sterile neutrinos can relieve the tension between the Planck\nobservation and the direct measurement of $H_0$ much better. Once we include\nthe $H_0$ measurement in the global fit, we find that the hint of the existence\nof sterile neutrinos in the holographic dark energy cosmology can be given.\nUnder the constraint of the all-data combination, we obtain $N_{\\rm eff}=\n3.76\\pm0.26$ and $m_{\\nu,\\rm sterile}^{\\rm eff}< 0.215\\,\\rm eV$, indicating\nthat the detection of $\\Delta N_{\\rm eff}>0$ in the holographic dark energy\ncosmology is at the $2.75\\sigma$ level and the massless or very light sterile\nneutrino is favored by the current observations.\n",
"title": "Search for sterile neutrinos in holographic dark energy cosmology: Reconciling Planck observation with the local measurement of the Hubble constant"
} | null | null | null | null | true | null | 20682 | null | Default | null | null |
null | {
"abstract": " In this work, we present a parallel, fully-distributed finite element\nnumerical framework to simulate the low-frequency electromagnetic response of\nsuperconducting devices, which allows to efficiently exploit HPC platforms. We\nselect the so-called H-formulation, which uses the magnetic field as a state\nvariable. Nédélec elements (of arbitrary order) are required for an\naccurate approximation of the H-formulation for modelling electromagnetic\nfields along interfaces between regions with high contrast medium properties.\nAn h-adaptive mesh refinement technique customized for Nédélec elements\nleads to a structured fine mesh in areas of interest whereas a smart coarsening\nis obtained in other regions. The composition of a tailored, robust, parallel\nnonlinear solver completes the exposition of the developed tools to tackle the\nproblem. First, a comparison against experimental data is performed to show the\navailability of the finite element approximation to model the physical\nphenomena. Then, a selected state-of-the-art 3D benchmark is reproduced,\nfocusing on the parallel performance of the algorithms.\n",
"title": "Simulation of high temperature superconductors and experimental validation"
} | null | null | null | null | true | null | 20683 | null | Default | null | null |
null | {
"abstract": " To improve system performance, modern operating systems (OSes) often\nundertake activities that require modification of virtual-to-physical page\ntranslation mappings. For example, the OS may migrate data between physical\nframes to defragment memory and enable superpages. The OS may migrate pages of\ndata between heterogeneous memory devices. We refer to all such activities as\npage remappings. Unfortunately, page remappings are expensive. We show that\ntranslation coherence is a major culprit and that systems employing\nvirtualization are especially badly affected by their overheads. In response,\nwe propose hardware translation invalidation and coherence or HATRIC, a readily\nimplementable hardware mechanism to piggyback translation coherence atop\nexisting cache coherence protocols. We perform detailed studies using KVM-based\nvirtualization, showing that HATRIC achieves up to 30% performance and 10%\nenergy benefits, for per-CPU area overheads of 2%. We also quantify HATRIC's\nbenefits on systems running Xen and find up to 33% performance improvements.\n",
"title": "Hardware Translation Coherence for Virtualized Systems"
} | null | null | null | null | true | null | 20684 | null | Default | null | null |
null | {
"abstract": " The interaction between proteins and DNA is a key driving force in a\nsignificant number of biological processes such as transcriptional regulation,\nrepair, recombination, splicing, and DNA modification. The identification of\nDNA-binding sites and the specificity of target proteins in binding to these\nregions are two important steps in understanding the mechanisms of these\nbiological activities. A number of high-throughput technologies have recently\nemerged that try to quantify the affinity between proteins and DNA motifs.\nDespite their success, these technologies have their own limitations and fall\nshort in precise characterization of motifs, and as a result, require further\ndownstream analysis to extract useful and interpretable information from a\nhaystack of noisy and inaccurate data. Here we propose MotifMark, a new\nalgorithm based on graph theory and machine learning, that can find binding\nsites on candidate probes and rank their specificity in regard to the\nunderlying transcription factor. We developed a pipeline to analyze\nexperimental data derived from compact universal protein binding microarrays\nand benchmarked it against two of the most accurate motif search methods. Our\nresults indicate that MotifMark can be a viable alternative technique for\nprediction of motif from protein binding microarrays and possibly other related\nhigh-throughput techniques.\n",
"title": "MotifMark: Finding Regulatory Motifs in DNA Sequences"
} | null | null | null | null | true | null | 20685 | null | Default | null | null |
null | {
"abstract": " This paper presents KeypointNet, an end-to-end geometric reasoning framework\nto learn an optimal set of category-specific 3D keypoints, along with their\ndetectors. Given a single image, KeypointNet extracts 3D keypoints that are\noptimized for a downstream task. We demonstrate this framework on 3D pose\nestimation by proposing a differentiable objective that seeks the optimal set\nof keypoints for recovering the relative pose between two views of an object.\nOur model discovers geometrically and semantically consistent keypoints across\nviewing angles and instances of an object category. Importantly, we find that\nour end-to-end framework using no ground-truth keypoint annotations outperforms\na fully supervised baseline using the same neural network architecture on the\ntask of pose estimation. The discovered 3D keypoints on the car, chair, and\nplane categories of ShapeNet are visualized at this http URL.\n",
"title": "Discovery of Latent 3D Keypoints via End-to-end Geometric Reasoning"
} | null | null | null | null | true | null | 20686 | null | Default | null | null |
null | {
"abstract": " A general theoretical framework is derived for the recently developed\nmulti-state trajectory (MST) approach from the time dependent Schrödinger\nequation, resulting in equations of motion for coupled nuclear-electronic\ndynamics equivalent to Hamilton dynamics or Heisenberg equation based on a new\nmultistate Meyer-Miller (MM) model. The derived MST formalism incorporates both\ndiabatic and adiabatic representations as limiting cases, and reduces to\nEhrenfest or Born-Oppenheimer dynamics in the mean field or the single state\nlimits, respectively. By quantizing nuclear dynamics to a particular active\nstate, the MST algorithm does not suffer from the instability caused by the\nnegative instant electronic population variables unlike the standard MM\ndynamics. Furthermore the multistate representation for electron coupled\nnuclear dynamics with each state associated with one individual trajectory\npresumably captures single state dynamics better than the mean field\ndescription. The coupled electronic-nuclear coherence is incorporated\nconsistently in the MST framework with no ad-hoc state switch and the\nassociated momentum adjustment or parameters for artificial decoherence, unlike\nthe original or modified surface hopping treatments. The implementation of the\nMST approach to benchmark problems shows reasonably good agreement with exact\nquantum calculations, and the results in both representations are similar in\naccuracy. The active state trajectory (AST) approximation of the MST approach\nprovides a consistent interpretation to trajectory surface hopping, which\npredicts the transition probabilities reasonably well for multiple nonadiabatic\ntransitions and conical intersection problems.\n",
"title": "Multi-State Trajectory Approach to Non-Adiabatic Dynamics: General Formalism and the Active State Trajectory Approximation"
} | null | null | null | null | true | null | 20687 | null | Default | null | null |
null | {
"abstract": " This paper presents a systematic approach for computing local solutions to\nmotion planning problems in non-convex environments using numerical optimal\ncontrol techniques. It extends the range of use of state-of-the-art numerical\noptimal control tools to problem classes where these tools have previously not\nbeen applicable. Today these problems are typically solved using motion\nplanners based on randomized or graph search. The general principle is to\ndefine a homotopy that perturbs, or preferably relaxes, the original problem to\nan easily solved problem. By combining a Sequential Quadratic Programming (SQP)\nmethod with a homotopy approach that gradually transforms the problem from a\nrelaxed one to the original one, practically relevant locally optimal solutions\nto the motion planning problem can be computed. The approach is demonstrated in\nmotion planning problems in challenging 2D and 3D environments, where the\npresented method significantly outperforms a state-of-the-art open-source\noptimizing sampled-based planner commonly used as benchmark.\n",
"title": "Combining Homotopy Methods and Numerical Optimal Control to Solve Motion Planning Problems"
} | null | null | [
"Mathematics"
]
| null | true | null | 20688 | null | Validated | null | null |
null | {
"abstract": " Atomically thin PtSe2 films have attracted extensive research interests for\npotential applications in high-speed electronics, spintronics and\nphotodetectors. Obtaining high quality, single crystalline thin films with\nlarge size is critical. Here we report the first successful layer-by-layer\ngrowth of high quality PtSe2 films by molecular beam epitaxy. Atomically thin\nfilms from 1 ML to 22 ML have been grown and characterized by low-energy\nelectron diffraction, Raman spectroscopy and X-ray photoemission spectroscopy.\nMoreover, a systematic thickness dependent study of the electronic structure is\nrevealed by angle-resolved photoemission spectroscopy (ARPES), and helical spin\ntexture is revealed by spin-ARPES. Our work provides new opportunities for\ngrowing large size single crystalline films for investigating the physical\nproperties and potential applications of PtSe2.\n",
"title": "High quality atomically thin PtSe2 films grown by molecular beam epitaxy"
} | null | null | null | null | true | null | 20689 | null | Default | null | null |
null | {
"abstract": " We present NAVREN-RL, an approach to NAVigate an unmanned aerial vehicle in\nan indoor Real ENvironment via end-to-end reinforcement learning RL. A suitable\nreward function is designed keeping in mind the cost and weight constraints for\nmicro drone with minimum number of sensing modalities. Collection of small\nnumber of expert data and knowledge based data aggregation is integrated into\nthe RL process to aid convergence. Experimentation is carried out on a Parrot\nAR drone in different indoor arenas and the results are compared with other\nbaseline technologies. We demonstrate how the drone successfully avoids\nobstacles and navigates across different arenas.\n",
"title": "NAVREN-RL: Learning to fly in real environment via end-to-end deep reinforcement learning using monocular images"
} | null | null | null | null | true | null | 20690 | null | Default | null | null |
null | {
"abstract": " Digital image correlation (DIC) is a widely used optical metrology for\nsurface deformation measurements. DIC relies on nonlinear optimization method.\nThus an initial guess is quite important due to its influence on the converge\ncharacteristics of the algorithm. In order to obtain a reliable, accurate\ninitial guess, a reliability-guided digital image correlation (RG-DIC) method,\nwhich is able to intelligently obtain a reliable initial guess without using\ntime-consuming integer-pixel registration, was proposed. However, the RG-DIC\nand its improved methods are path-dependent and cannot be fully parallelized.\nBesides, it is highly possible that RG-DIC fails in the full-field analysis of\ndeformation without manual intervention if the deformation fields contain large\nareas of discontinuous deformation. Feature-based initial guess is highly\nrobust while it is relatively time-consuming. Recently, path-independent\nalgorithm, fast Fourier transform-based cross correlation (FFT-CC) algorithm,\nwas proposed to estimate the initial guess. Complete parallelizability is the\nmajor advantage of the FFT-CC algorithm, while it is sensitive to small\ndeformation. Wu et al proposed an efficient integer-pixel search scheme, but\nthe parameters of this algorithm are set by the users empirically. In this\ntechnical note, a fully parallelizable DIC method is proposed. Different from\nRG-DIC method, the proposed method divides DIC algorithm into two parts:\nfull-field initial guess estimation and sub-pixel registration. The proposed\nmethod has the following benefits: 1) providing a pre-knowledge of deformation\nfields; 2) saving computational time; 3) reducing error propagation; 4)\nintegratability with well-established DIC algorithms; 5) fully\nparallelizability.\n",
"title": "Fast, Accurate and Fully Parallelizable Digital Image Correlation"
} | null | null | null | null | true | null | 20691 | null | Default | null | null |
null | {
"abstract": " The classical idea of evolutionarily stable strategy (ESS) modeling animal\nbehavior does not involve any spatial dependence. We considered a spatial\nHawk-Dove game played by animals in a patchy environment with wrap around\nboundaries. We posit that each site contains the same number of individuals. An\nevolution equation for analyzing the stability of the ESS is found as the mean\ndynamics of the classical frequency dependent Moran process coupled via\nmigration and nonlocal payoff calculation in 1D and 2D habitats. The linear\nstability analysis of the model is performed and conditions to observe spatial\npatterns are investigated. For the nearest neighbor interactions (including von\nNeumann and Moore neighborhoods in 2D) we concluded that it is possible to\ndestabilize the ESS of the game and observe pattern formation when the\ndispersal rate is small enough. We numerically investigate the spatial patterns\narising from the replicator equations coupled via nearest neighbor payoff\ncalculation and dispersal.\n",
"title": "Discovering the effect of nonlocal payoff calculation on the stabilty of ESS: Spatial patterns of Hawk-Dove game in metapopulations"
} | null | null | null | null | true | null | 20692 | null | Default | null | null |
null | {
"abstract": " Measurements of the high-frequency complex resistivity in superconductors are\na tool often used to obtain the vortex parameters, such as the vortex\nviscosity, the pinning constant and the depinning frequency. In anisotropic\nsuperconductors, the extraction of these quantities from the measurements faces\nnew difficulties due to the tensor nature of the electromagnetic problem. The\nproblem is specifically intricate when the magnetic field is tilted with\nrespect to the crystallographic axes. Partial solutions exist in the\nfree-flux-flow (no pinning) and Campbell (pinning dominated) regimes. In this\npaper we develop a full tensor model for the vortex motion complex resistivity,\nincluding flux-flow, pinning, and creep. We give explicit expressions for the\ntensors involved. We obtain that, despite the complexity of the physics, some\nparameters remain scalar in nature. We show that under specific circumstances\nthe directly measured quantities do not reflect the true vortex parameters, and\nwe give procedures to derive the true vortex parameters from measurements taken\nwith arbitrary field orientations. Finally, we discuss the applicability of the\nangular scaling properties to the measured and transformed vortex parameters\nand we exploit these properties as a tool to unveil the existence of\ndirectional pinning.\n",
"title": "Analysis of the measurements of anisotropic a.c. vortex resistivity in tilted magnetic fields"
} | null | null | null | null | true | null | 20693 | null | Default | null | null |
null | {
"abstract": " This paper presents an empirical study on applying convolutional neural\nnetworks (CNNs) to detecting J-UNIWARD, one of the most secure JPEG\nsteganographic method. Experiments guiding the architectural design of the CNNs\nhave been conducted on the JPEG compressed BOSSBase containing 10,000 covers of\nsize 512x512. Results have verified that both the pooling method and the depth\nof the CNNs are critical for performance. Results have also proved that a\n20-layer CNN, in general, outperforms the most sophisticated feature-based\nmethods, but its advantage gradually diminishes on hard-to-detect cases. To\nshow that the performance generalizes to large-scale databases and to different\ncover sizes, one experiment has been conducted on the CLS-LOC dataset of\nImageNet containing more than one million covers cropped to unified size of\n256x256. The proposed 20-layer CNN has cut the error achieved by a CNN recently\nproposed for large-scale JPEG steganalysis by 35%. Source code is available via\nGitHub: this https URL\n",
"title": "Deep Convolutional Neural Network to Detect J-UNIWARD"
} | null | null | [
"Computer Science"
]
| null | true | null | 20694 | null | Validated | null | null |
null | {
"abstract": " In this letter we present a measurement of the phase-space density\ndistribution (PSDD) of ultra-cold \\Rb atoms performing 1D anomalous diffusion.\nThe PSDD is imaged using a direct tomographic method based on Raman velocity\nselection. It reveals that the position-velocity correlation function\n$C_{xv}(t)$ builds up on a timescale related to the initial conditions of the\nensemble and then decays asymptotically as a power-law. We show that the decay\nfollows a simple scaling theory involving the power-law asymptotic dynamics of\nposition and velocity. The generality of this scaling theory is confirmed using\nMonte-Carlo simulations of two distinct models of anomalous diffusion.\n",
"title": "Observing Power-Law Dynamics of Position-Velocity Correlation in Anomalous Diffusion"
} | null | null | [
"Physics"
]
| null | true | null | 20695 | null | Validated | null | null |
null | {
"abstract": " The $E_8$ root lattice can be constructed from the modular curve $X(13)$ by\nthe invariant theory for the simple group $\\text{PSL}(2, 13)$. This gives a\ndifferent construction of the $E_8$ root lattice. It also gives an explicit\nconstruction of the modular curve $X(13)$.\n",
"title": "Modular curves, invariant theory and $E_8$"
} | null | null | [
"Mathematics"
]
| null | true | null | 20696 | null | Validated | null | null |
null | {
"abstract": " We present convergence rate analysis for the approximate stochastic gradient\nmethod, where individual gradient updates are corrupted by computation errors.\nWe develop stochastic quadratic constraints to formulate a small linear matrix\ninequality (LMI) whose feasible set characterizes convergence properties of the\napproximate stochastic gradient. Based on this LMI condition, we develop a\nsequential minimization approach to analyze the intricate trade-offs that\ncouple stepsize selection, convergence rate, optimization accuracy, and\nrobustness to gradient inaccuracy. We also analytically solve this LMI\ncondition and obtain theoretical formulas that quantify the convergence\nproperties of the approximate stochastic gradient under various assumptions on\nthe loss functions.\n",
"title": "Analysis of Approximate Stochastic Gradient Using Quadratic Constraints and Sequential Semidefinite Programs"
} | null | null | null | null | true | null | 20697 | null | Default | null | null |
null | {
"abstract": " We give a description of complex geodesics and we study the structure of\nstationary discs in some non-convex domains for which complex geodesics are not\nunique.\n",
"title": "Invariant holomorphic discs in some non-convex domains"
} | null | null | null | null | true | null | 20698 | null | Default | null | null |
null | {
"abstract": " Recent advances in bioinformatics have made high-throughput microbiome data\nwidely available, and new statistical tools are required to maximize the\ninformation gained from these data. For example, analysis of high-dimensional\nmicrobiome data from designed experiments remains an open area in microbiome\nresearch. Contemporary analyses work on metrics that summarize collective\nproperties of the microbiome, but such reductions preclude inference on the\nfine-scale effects of environmental stimuli on individual microbial taxa. Other\napproaches model the proportions or counts of individual taxa as response\nvariables in mixed models, but these methods fail to account for complex\ncorrelation patterns among microbial communities. In this paper, we propose a\nnovel Bayesian mixed-effects model that exploits cross-taxa correlations within\nthe microbiome, a model we call MIMIX (MIcrobiome MIXed model). MIMIX offers\nglobal tests for treatment effects, local tests and estimation of treatment\neffects on individual taxa, quantification of the relative contribution from\nheterogeneous sources to microbiome variability, and identification of latent\necological subcommunities in the microbiome. MIMIX is tailored to large\nmicrobiome experiments using a combination of Bayesian factor analysis to\nefficiently represent dependence between taxa and Bayesian variable selection\nmethods to achieve sparsity. We demonstrate the model using a simulation\nexperiment and on a 2x2 factorial experiment of the effects of nutrient\nsupplement and herbivore exclusion on the foliar fungal microbiome of\n$\\textit{Andropogon gerardii}$, a perennial bunchgrass, as part of the global\nNutrient Network research initiative.\n",
"title": "MIMIX: a Bayesian Mixed-Effects Model for Microbiome Data from Designed Experiments"
} | null | null | null | null | true | null | 20699 | null | Default | null | null |
null | {
"abstract": " This article illustrates how to measure the heterogeneity of spatial data\npresenting a finite number of categories via computation of spatial entropy.\nThe R package SpatEntropy contains functions for the computation of entropy and\nspatial entropy measures. The extension to spatial entropy measures is a unique\nfeature of SpatEntropy. In addition to the traditional version of Shannon's\nentropy, the package includes Batty's spatial entropy, O'Neill's entropy, Li\nand Reynolds' contagion index, Karlstrom and Ceccato's entropy, Leibovici's\nentropy, Parresol and Edwards' entropy and Altieri's entropy. The package is\nable to work with both areal and point data. This paper is a general\ndescription of SpatEntropy, as well as its necessary theoretical background,\nand an introduction for new users.\n",
"title": "SpatEntropy: Spatial Entropy Measures in R"
} | null | null | null | null | true | null | 20700 | null | Default | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.