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": " A major challenge in computational chemistry is the generation of novel\nmolecular structures with desirable pharmacological and physiochemical\nproperties. In this work, we investigate the potential use of autoencoder, a\ndeep learning methodology, for de novo molecular design. Various generative\nautoencoders were used to map molecule structures into a continuous latent\nspace and vice versa and their performance as structure generator was assessed.\nOur results show that the latent space preserves chemical similarity principle\nand thus can be used for the generation of analogue structures. Furthermore,\nthe latent space created by autoencoders were searched systematically to\ngenerate novel compounds with predicted activity against dopamine receptor type\n2 and compounds similar to known active compounds not included in the training\nset were identified.\n",
"title": "Application of generative autoencoder in de novo molecular design"
} | null | null | null | null | true | null | 19801 | null | Default | null | null |
null | {
"abstract": " Measurements of cosmic microwave background spectral distortions have\nprofound implications for our understanding of physical processes taking place\nover a vast window in cosmological history. Foreground contamination is\nunavoidable in such measurements and detailed signal-foreground separation will\nbe necessary to extract cosmological science. We present MCMC-based spectral\ndistortion detection forecasts in the presence of Galactic and extragalactic\nforegrounds for a range of possible experimental configurations, focusing on\nthe Primordial Inflation Explorer (PIXIE) as a fiducial concept. We consider\nmodifications to the baseline PIXIE mission (operating 12 months in distortion\nmode), searching for optimal configurations using a Fisher approach. Using only\nspectral information, we forecast an extended PIXIE mission to detect the\nexpected average non-relativistic and relativistic thermal Sunyaev-Zeldovich\ndistortions at high significance (194$\\sigma$ and 11$\\sigma$, respectively),\neven in the presence of foregrounds. The $\\Lambda$CDM Silk damping $\\mu$-type\ndistortion is not detected without additional modifications of the instrument\nor external data. Galactic synchrotron radiation is the most problematic source\nof contamination in this respect, an issue that could be mitigated by combining\nPIXIE data with future ground-based observations at low frequencies ($\\nu <\n15-30$GHz). Assuming moderate external information on the synchrotron spectrum,\nwe project an upper limit of $|\\mu| < 3.6\\times 10^{-7}$ (95\\% c.l.), slightly\nmore than one order of magnitude above the fiducial $\\Lambda$CDM signal from\nthe damping of small-scale primordial fluctuations, but a factor of $\\simeq\n250$ improvement over the current upper limit from COBE/FIRAS. This limit could\nbe further reduced to $|\\mu| < 9.4\\times 10^{-8}$ (95\\% c.l.) with more\noptimistic assumptions about low-frequency information. (Abridged)\n",
"title": "Prospects for Measuring Cosmic Microwave Background Spectral Distortions in the Presence of Foregrounds"
} | null | null | null | null | true | null | 19802 | null | Default | null | null |
null | {
"abstract": " For over a decade, scientists at NASA's Jet Propulsion Laboratory (JPL) have\nbeen recording measurements from the Martian surface as a part of the Mars\nExploration Rovers mission. One quantity of interest has been the opacity of\nMars's atmosphere for its importance in day-to-day estimations of the amount of\npower available to the rover from its solar arrays. This paper proposes the use\nof neural networks as a method for forecasting Martian atmospheric opacity that\nis more effective than the current empirical model. The more accurate\nprediction provided by these networks would allow operators at JPL to make more\naccurate predictions of the amount of energy available to the rover when they\nplan activities for coming sols.\n",
"title": "The Fog of War: A Machine Learning Approach to Forecasting Weather on Mars"
} | null | null | null | null | true | null | 19803 | null | Default | null | null |
null | {
"abstract": " Proteins are only moderately stable. It has long been debated whether this\nnarrow range of stabilities is solely a result of neutral drift towards lower\nstability or purifying selection against excess stability is also at work - for\nwhich no experimental evidence was found so far. Here we show that mutations\noutside the active site in the essential E. coli enzyme adenylate kinase result\nin stability-dependent increase in substrate inhibition by AMP, thereby\nimpairing overall enzyme activity at high stability. Such inhibition caused\nsubstantial fitness defects not only in the presence of excess substrate but\nalso under physiological conditions. In the latter case, substrate inhibition\ncaused differential accumulation of AMP in the stationary phase for the\ninhibition prone mutants. Further, we show that changes in flux through Adk\ncould accurately describe the variation in fitness effects. Taken together,\nthese data suggest that selection against substrate inhibition and hence excess\nstability may have resulted in a narrow range of optimal stability observed for\nmodern proteins.\n",
"title": "Substrate inhibition imposes fitness penalty at high protein stability"
} | null | null | null | null | true | null | 19804 | null | Default | null | null |
null | {
"abstract": " We provide a proposal, motivated by Separation of Variables and gauge theory\narguments, for constructing exact solutions to the quantum Baxter equation\nassociated to the $N$-particle relativistic Toda chain and test our proposal\nagainst numerical results. Quantum Mechanical non-perturbative corrections,\nessential in order to obtain a sensible solution, are taken into account in our\ngauge theory approach by considering codimension two defects on curved\nbackgrounds (squashed $S^5$ and degenerate limits) rather than flat space; this\nsetting also naturally incorporates exact quantization conditions and energy\nspectrum of the relativistic Toda chain as well as its modular dual structure.\n",
"title": "Exact relativistic Toda chain eigenfunctions from Separation of Variables and gauge theory"
} | null | null | null | null | true | null | 19805 | null | Default | null | null |
null | {
"abstract": " A monotone drawing of a graph G is a straight-line drawing of G such that\nevery pair of vertices is connected by a path that is monotone with respect to\nsome direction.\nTrees, as a special class of graphs, have been the focus of several papers\nand, recently, He and He~\\cite{mt:4} showed how to produce a monotone drawing\nof an arbitrary $n$-vertex tree that is contained in a $12n \\times 12n$ grid.\nAll monotone tree drawing algorithms that have appeared in the literature\nconsider rooted ordered trees and they draw them so that (i) the root of the\ntree is drawn at the origin of the drawing, (ii) the drawing is confined in the\nfirst quadrant, and (iii) the ordering/embedding of the tree is respected. In\nthis paper, we provide a simple algorithm that has the exact same\ncharacteristics and, given an $n$-vertex rooted tree $T$, it outputs a monotone\ndrawing of $T$ that fits on a $n \\times n$ grid.\nFor unrooted ordered trees, we present an algorithms that produces monotone\ndrawings that respect the ordering and fit in an $(n+1) \\times (\\frac{n}{2}\n+1)$ grid, while, for unrooted non-ordered trees we produce monotone drawings\nof good aspect ratio which fit on a grid of size at most $\\left\\lfloor\n\\frac{3}{4} \\left(n+2\\right)\\right\\rfloor \\times \\left\\lfloor \\frac{3}{4}\n\\left(n+2\\right)\\right\\rfloor$.\n",
"title": "Simple Compact Monotone Tree Drawings"
} | null | null | null | null | true | null | 19806 | null | Default | null | null |
null | {
"abstract": " Graph data models are widely used in many areas, for example, bioinformatics,\ngraph databases. In these areas, it is often required to process queries for\nlarge graphs. Some of the most common graph queries are navigational queries.\nThe result of query evaluation is a set of implicit relations between nodes of\nthe graph, i.e. paths in the graph. A natural way to specify these relations is\nby specifying paths using formal grammars over the alphabet of edge labels. An\nanswer to a context-free path query in this approach is usually a set of\ntriples (A, m, n) such that there is a path from the node m to the node n,\nwhose labeling is derived from a non-terminal A of the given context-free\ngrammar. This type of queries is evaluated using the relational query\nsemantics. Another example of path query semantics is the single-path query\nsemantics which requires presenting a single path from the node m to the node\nn, whose labeling is derived from a non-terminal A for all triples (A, m, n)\nevaluated using the relational query semantics. There is a number of algorithms\nfor query evaluation which use these semantics but all of them perform poorly\non large graphs. One of the most common technique for efficient big data\nprocessing is the use of a graphics processing unit (GPU) to perform\ncomputations, but these algorithms do not allow to use this technique\nefficiently. In this paper, we show how the context-free path query evaluation\nusing these query semantics can be reduced to the calculation of the matrix\ntransitive closure. Also, we propose an algorithm for context-free path query\nevaluation which uses relational query semantics and is based on matrix\noperations that make it possible to speed up computations by using a GPU.\n",
"title": "Context-Free Path Querying by Matrix Multiplication"
} | null | null | null | null | true | null | 19807 | null | Default | null | null |
null | {
"abstract": " We remark that sparse and Carleson coefficients are equivalent for every\ncountable collection of Borel sets and hence, in particular, for dyadic\nrectangles, the case relevant to the theory of bi-parameter singular integrals.\nThe key observation is that a dual refomulation by I. E. Verbitsky for\nCarleson coefficients over dyadic cubes holds also for Carleson coefficients\nover general sets. We give a simple proof for this reformulation.\n",
"title": "Equivalence of sparse and Carleson coefficients for general sets"
} | null | null | [
"Mathematics"
]
| null | true | null | 19808 | null | Validated | null | null |
null | {
"abstract": " Discovering business rules from business process models are of advantage to\nensure the compliance of business processes with business rules. Furthermore it\nprovides the agility of business processes in case of business rules evolution.\nCurrent approaches are limited on types of rules that can be discovered. This\npaper analyses the expression power of some popular business process modelling\nlanguages in embedding business rules in its presentation and provides\nindicators to extract various types of business rules from business process\nmodels.\n",
"title": "Discovering Business Rules from Business Process Models"
} | null | null | [
"Computer Science"
]
| null | true | null | 19809 | null | Validated | null | null |
null | {
"abstract": " We combine features extracted from pre-trained convolutional neural networks\n(CNNs) with the fast, linear Exemplar-LDA classifier to get the advantages of\nboth: the high detection performance of CNNs, automatic feature engineering,\nfast model learning from few training samples and efficient sliding-window\ndetection. The Adaptive Real-Time Object Detection System (ARTOS) has been\nrefactored broadly to be used in combination with Caffe for the experimental\nstudies reported in this work.\n",
"title": "Fast Learning and Prediction for Object Detection using Whitened CNN Features"
} | null | null | null | null | true | null | 19810 | null | Default | null | null |
null | {
"abstract": " Context. Planet formation with pebbles has been proposed to solve a couple of\nlong-standing issues in the classical formation model. Some sophisticated\nsimulations have been done to confirm the efficiency of pebble accretion.\nHowever, there has not been any global N-body simulations that compare the\noutcomes of planet formation via pebble accretion with observed extrasolar\nplanetary systems. Aims. In this paper, we study the effects of a range of\ninitial parameters of planet formation via pebble accretion, and present the\nfirst results of our simulations. Methods. We incorporate the pebble accretion\nmodel by Ida et al. (2016) in the N-body code SyMBA (Duncan et al. 1998), along\nwith the effects of gas accretion, eccentricity and inclination damping and\nplanet migration in the disc. Results. We confirm that pebble accretion leads\nto a variety of planetary systems, but have difficulty in reproducing observed\nproperties of exoplanetary systems, such as planetary mass, semimajor axis, and\neccentricity distributions. The main reason behind this is a too-efficient type\nI migration, which sensitively depends on the disc model. However, our\nsimulations also lead to a few interesting predictions. First, we find that\nformation efficiencies of planets depend on the stellar metallicities, not only\nfor giant planets, but also for Earths (Es) and Super-Earths (SEs). The\ndependency for Es/SEs is subtle. Although higher metallicity environments lead\nto faster formation of a larger number of Es/SEs, they also tend to be lost\nlater via dynamical instability. Second, our results indicate that a wide range\nof bulk densities observed for Es and SEs is a natural consequence of dynamical\nevolution of planetary systems. Third, the ejection trend of our simulations\nsuggest that one free-floating E/SE may be expected for two smaller-mass\nplanets.\n",
"title": "N-body simulations of planet formation via pebble accretion I: First Results"
} | null | null | [
"Physics"
]
| null | true | null | 19811 | null | Validated | null | null |
null | {
"abstract": " We show that NSOP$_{1}$ theories are exactly the theories in which\nKim-independence satisfies a form of local character. In particular, we show\nthat if $T$ is NSOP$_{1}$, $M\\models T$, and $p$ is a type over $M$, then the\ncollection of elementary substructures of size $\\left|T\\right|$ over which $p$\ndoes not Kim-fork is a club of $\\left[M\\right]^{\\left|T\\right|}$ and that this\ncharacterizes NSOP$_{1}$.\nWe also present a new phenomenon we call dual local-character for\nKim-independence in NSOP$_{1}$-theories.\n",
"title": "Local character of Kim-independence"
} | null | null | null | null | true | null | 19812 | null | Default | null | null |
null | {
"abstract": " Heterogeneity has been studied as one of the most common explanations of the\npuzzle of cooperation in social dilemmas. A large number of papers have been\npublished discussing the effects of increasing heterogeneity in structured\npopulations of agents, where it has been established that heterogeneity may\nfavour cooperative behaviour if it supports agents to locally coordinate their\nstrategies. In this paper, assuming an existing model of a heterogeneous\nweighted network, we aim to further this analysis by exploring the relationship\n(if any) between heterogeneity and cooperation. We adopt a weighted network\nwhich is fully populated by agents playing both the Prisoner's Dilemma or the\nOptional Prisoner's Dilemma games with coevolutionary rules, i.e., not only the\nstrategies but also the link weights evolve over time. Surprisingly, results\nshow that the heterogeneity of link weights (states) on their own does not\nalways promote cooperation; rather cooperation is actually favoured by the\nincrease in the number of overlapping states and not by the heterogeneity\nitself. We believe that these results can guide further research towards a more\naccurate analysis of the role of heterogeneity in social dilemmas.\n",
"title": "A Further Analysis of The Role of Heterogeneity in Coevolutionary Spatial Games"
} | null | null | null | null | true | null | 19813 | null | Default | null | null |
null | {
"abstract": " The magnetorotational instability (MRI) is thought to be a powerful source of\nturbulence in Keplerian accretion disks. Motivated by recent laboratory\nexperiments, we study the MRI driven by an azimuthal magnetic field in an\nelectrically conducting fluid sheared between two concentric rotating\ncylinders. By adjusting the rotation rates of the cylinders, we approximate\nangular velocity profiles $\\omega \\propto r^{q}$. We perform direct numerical\nsimulations of a steep profile close to the Rayleigh line $q \\gtrsim -2 $ and a\nquasi-Keplerian profile $q \\approx -3/2$ and cover wide ranges of Reynolds\n($Re\\le 4\\cdot10^4$) and magnetic Prandtl numbers ($0\\le Pm \\le 1$). In the\nquasi-Keplerian case, the onset of instability depends on the magnetic Reynolds\nnumber, with $Rm_c \\approx 50$, and angular momentum transport scales as\n$\\sqrt{Pm} Re^2$ in the turbulent regime. The ratio of Maxwell to Reynolds\nstresses is set by $Rm$. At the onset of instability both stresses have similar\nmagnitude, whereas the Reynolds stress vanishes or becomes even negative as\n$Rm$ increases. For the profile close to the Rayleigh line, the instability\nshares these properties as long as $Pm\\gtrsim0.1$, but exhibits a markedly\ndifferent character if $Pm\\rightarrow 0$, where the onset of instability is\ngoverned by the Reynolds number, with $Re_c \\approx 1250$, transport is via\nReynolds stresses and scales as $Re^2$. At intermediate $Pm=0.01$ we observe a\ncontinuous transition from one regime to the other, with a crossover at\n$Rm=\\mathcal{O}(100)$. Our results give a comprehensive picture of angular\nmomentum transport of the MRI with an imposed azimuthal field.\n",
"title": "Transport properties of the Azimuthal Magnetorotational Instability"
} | null | null | null | null | true | null | 19814 | null | Default | null | null |
null | {
"abstract": " An individual has been subjected to some exposure and has developed some\noutcome. Using data on similar individuals, we wish to evaluate, for this case,\nthe probability that the outcome was in fact caused by the exposure. Even with\nthe best possible experimental data on exposure and outcome, we typically can\nnot identify this \"probability of causation\" exactly, but we can provide\ninformation in the form of bounds for it. Under appropriate assumptions, these\nbounds can be tightened if we can make other observations (e.g., on\nnon-experimental cases), measure additional variables (e.g., covariates) or\nmeasure complete mediators. In this work we propose new bounds for the case\nthat a third variable mediates partially the effect of the exposure on the\noutcome.\n",
"title": "New bounds for the Probability of Causation in Mediation Analysis"
} | null | null | [
"Mathematics",
"Statistics"
]
| null | true | null | 19815 | null | Validated | null | null |
null | {
"abstract": " It is well known that the Milgrom's MOND (modified Newtonian dynamics)\nexplains well the mass discrepancy problem in galaxy rotation curves. The MOND\npredicts a universal acceleration scale below which the Newtonian dynamics is\ninvalid yet. The universal acceleration scale we got from the SPARC dataset is\n$g_†=1.02\\times10^{-10} \\rm m~s^{-2}$. Milgrom suggested that the\nacceleration scale may be a fingerprint of cosmology on local dynamics and\nrelated with the Hubble constant $g_†\\sim cH_0$. In this paper, we use the\nhemisphere comparison method with the SPARC dataset to investigate the spatial\nanisotropy on the acceleration scale. We find that the hemisphere of the\nmaximum acceleration scale is in the direction $(l,b) =\n({175.5^\\circ}^{+6^\\circ}_{-10^\\circ}, {-6.5^\\circ}^{+8^\\circ}_{-3^\\circ})$\nwith $g_{†,max}=1.10\\times10^{-10} \\rm m~s^{-2}$, while the hemisphere of\nthe minimum acceleration scale is in the opposite direction $(l,b) =\n({355.5^\\circ}^{+6^\\circ}_{-10^\\circ}, {6.5^\\circ}^{+3^\\circ}_{-8^\\circ})$ with\n$g_{†,min}=0.76\\times10^{-10} \\rm m~s^{-2}$. The maximum anisotropy level\nreaches up to $0.37\\pm0.04$. Robust tests present that such a level of\nanisotropy can't be reproduced by a statistically isotropic data. In addition,\nwe show that the spatial anisotropy on the acceleration scale has little\ncorrelation with the non-uniform distribution of the SPARC data points in sky.\nWe also find that the maximum anisotropy direction is close with other\ncosmological preferred directions, especially the direction of the \"Australia\ndipole\" for the fine structure constant.\n",
"title": "Searching for a cosmological preferred direction with 147 rotationally supported galaxies"
} | null | null | null | null | true | null | 19816 | null | Default | null | null |
null | {
"abstract": " We present a framework for testing independence between two random vectors\nthat is scalable to massive data. We break down the multivariate test into\nunivariate tests of independence on a collection of $2\\times 2$ contingency\ntables, constructed by sequentially discretizing the sample space. This\ntransforms a complex problem that traditionally requires quadratic\ncomputational complexity with respect to the sample size into one that scales\nalmost linearly with the sample size. We further consider the scenario when the\ndimensionality of the random vectors grows large, in which case the curse of\ndimensionality arises in the proposed framework through an explosion in the\nnumber of univariate tests to be completed. To overcome this difficulty we\npropose a data-adaptive version of our method that completes a fraction of the\nunivariate tests judged to be more likely to contain evidence for dependency.\nWe demonstrate the tremendous computational advantage of the algorithm in\ncomparison to existing approaches while achieving desirable statistical power\nthrough an extensive simulation study. In addition, we illustrate how our\nmethod can be used for learning the nature of the underlying dependency. We\ndemonstrate the use of our method through analyzing a data set from flow\ncytometry.\n",
"title": "MultiFIT: A Multivariate Multiscale Framework for Independence Tests"
} | null | null | null | null | true | null | 19817 | null | Default | null | null |
null | {
"abstract": " This is a survey article, based on the author's lectures in the 2015 AMS\nSummer Research Institute in Algebraic Geometry, and to appear in the\nProceedings.\n",
"title": "Stability of algebraic varieties and Kahler geometry"
} | null | null | null | null | true | null | 19818 | null | Default | null | null |
null | {
"abstract": " The surface of metal, glass and plastic objects is often characterized by\nmicroscopic scratches caused by manufacturing and/or wear. A closer look onto\nsuch scratches reveals iridescent colors with a complex dependency on viewing\nand lighting conditions. The physics behind this phenomenon is well understood;\nit is caused by diffraction of the incident light by surface features on the\norder of the optical wavelength. Existing analytic models are able to reproduce\nspatially unresolved microstructure such as the iridescent appearance of\ncompact disks and similar materials. Spatially resolved scratches, on the other\nhand, have proven elusive due to the highly complex wave-optical light\ntransport simulations needed to account for their appearance. In this paper, we\npropose a wave-optical shading model based on non-paraxial scalar diffraction\ntheory to render this class of effects. Our model expresses surface roughness\nas a collection of line segments. To shade a point on the surface, the\nindividual diffraction patterns for contributing scratch segments are computed\nanalytically and superimposed coherently. This provides natural transitions\nfrom localized glint-like iridescence to smooth BRDFs representing the\nsuperposition of many reflections at large viewing distances. We demonstrate\nthat our model is capable of recreating the overall appearance as well as\ncharacteristic detail effects observed on real-world examples.\n",
"title": "Scratch iridescence: Wave-optical rendering of diffractive surface structure"
} | null | null | null | null | true | null | 19819 | null | Default | null | null |
null | {
"abstract": " The oscillatory behavior of the solutions to a differential equation with\nseveral non-monotone delay arguments and non-negative coefficients is studied.\nA new sufficient oscillation condition, involving lim sup, is obtained. An\nexample illustrating the significance of the result is also given.\n",
"title": "An oscillation criterion for delay differential equations with several non-monotone arguments"
} | null | null | null | null | true | null | 19820 | null | Default | null | null |
null | {
"abstract": " In recent years, Variational Autoencoders (VAEs) have been shown to be highly\neffective in both standard collaborative filtering applications and extensions\nsuch as incorporation of implicit feedback. We extend VAEs to collaborative\nfiltering with side information, for instance when ratings are combined with\nexplicit text feedback from the user. Instead of using a user-agnostic standard\nGaussian prior, we incorporate user-dependent priors in the latent VAE space to\nencode users' preferences as functions of the review text. Taking into account\nboth the rating and the text information to represent users in this multimodal\nlatent space is promising to improve recommendation quality. Our proposed model\nis shown to outperform the existing VAE models for collaborative filtering (up\nto 29.41% relative improvement in ranking metric) along with other baselines\nthat incorporate both user ratings and text for item recommendation.\n",
"title": "Item Recommendation with Variational Autoencoders and Heterogenous Priors"
} | null | null | null | null | true | null | 19821 | null | Default | null | null |
null | {
"abstract": " Boltzmann exploration is a classic strategy for sequential decision-making\nunder uncertainty, and is one of the most standard tools in Reinforcement\nLearning (RL). Despite its widespread use, there is virtually no theoretical\nunderstanding about the limitations or the actual benefits of this exploration\nscheme. Does it drive exploration in a meaningful way? Is it prone to\nmisidentifying the optimal actions or spending too much time exploring the\nsuboptimal ones? What is the right tuning for the learning rate? In this paper,\nwe address several of these questions in the classic setup of stochastic\nmulti-armed bandits. One of our main results is showing that the Boltzmann\nexploration strategy with any monotone learning-rate sequence will induce\nsuboptimal behavior. As a remedy, we offer a simple non-monotone schedule that\nguarantees near-optimal performance, albeit only when given prior access to key\nproblem parameters that are typically not available in practical situations\n(like the time horizon $T$ and the suboptimality gap $\\Delta$). More\nimportantly, we propose a novel variant that uses different learning rates for\ndifferent arms, and achieves a distribution-dependent regret bound of order\n$\\frac{K\\log^2 T}{\\Delta}$ and a distribution-independent bound of order\n$\\sqrt{KT}\\log K$ without requiring such prior knowledge. To demonstrate the\nflexibility of our technique, we also propose a variant that guarantees the\nsame performance bounds even if the rewards are heavy-tailed.\n",
"title": "Boltzmann Exploration Done Right"
} | null | null | [
"Computer Science",
"Statistics"
]
| null | true | null | 19822 | null | Validated | null | null |
null | {
"abstract": " For sophisticated reinforcement learning (RL) systems to interact usefully\nwith real-world environments, we need to communicate complex goals to these\nsystems. In this work, we explore goals defined in terms of (non-expert) human\npreferences between pairs of trajectory segments. We show that this approach\ncan effectively solve complex RL tasks without access to the reward function,\nincluding Atari games and simulated robot locomotion, while providing feedback\non less than one percent of our agent's interactions with the environment. This\nreduces the cost of human oversight far enough that it can be practically\napplied to state-of-the-art RL systems. To demonstrate the flexibility of our\napproach, we show that we can successfully train complex novel behaviors with\nabout an hour of human time. These behaviors and environments are considerably\nmore complex than any that have been previously learned from human feedback.\n",
"title": "Deep reinforcement learning from human preferences"
} | null | null | null | null | true | null | 19823 | null | Default | null | null |
null | {
"abstract": " We present a quantum algorithm to compute the entanglement spectrum of\narbitrary quantum states. The interesting universal part of the entanglement\nspectrum is typically contained in the largest eigenvalues of the density\nmatrix which can be obtained from the lower Renyi entropies through the\nNewton-Girard method. Obtaining the $p$ largest eigenvalues\n($\\lambda_1>\\lambda_2\\ldots>\\lambda_p$) requires a parallel circuit depth of\n$\\mathcal{O}(p(\\lambda_1/\\lambda_p)^p)$ and $\\mathcal{O}(p\\log(N))$ qubits\nwhere up to $p$ copies of the quantum state defined on a Hilbert space of size\n$N$ are needed as the input. We validate this procedure for the entanglement\nspectrum of the topologically-ordered Laughlin wave function corresponding to\nthe quantum Hall state at filling factor $\\nu=1/3$. Our scaling analysis\nexposes the tradeoffs between time and number of qubits for obtaining the\nentanglement spectrum in the thermodynamic limit using finite-size digital\nquantum computers. We also illustrate the utility of the second Renyi entropy\nin predicting a topological phase transition and in extracting the localization\nlength in a many-body localized system.\n",
"title": "Entanglement spectroscopy on a quantum computer"
} | null | null | [
"Physics"
]
| null | true | null | 19824 | null | Validated | null | null |
null | {
"abstract": " The Jiangmen Underground Neutrino Observatory (JUNO) is a multipurpose\nneutrino experiment with a 20 kt liquid scintillator detector designed to\ndetermine the neutrino mass hierarchy, and measure the neutrino oscillation\nparameters. Linear alkyl benzene (LAB) will be used as the solvent for the\nliquid scintillation system in the central detector of JUNO. For this purpose,\nwe have prepared LAB samples, and have measured their light attenuation\nlengths, with one achieving a length of 25.8 m, comparable to the diameter of\nthe JUNO detector.\n",
"title": "Light Attenuation Length of High Quality Linear Alkyl Benzene as Liquid Scintillator Solvent for the JUNO Experiment"
} | null | null | null | null | true | null | 19825 | null | Default | null | null |
null | {
"abstract": " Recent work has explored the problem of autonomous navigation by imitating a\nteacher and learning an end-to-end policy, which directly predicts controls\nfrom raw images. However, these approaches tend to be sensitive to mistakes by\nthe teacher and do not scale well to other environments or vehicles. To this\nend, we propose Observational Imitation Learning (OIL), a novel imitation\nlearning variant that supports online training and automatic selection of\noptimal behavior by observing multiple imperfect teachers. We apply our\nproposed methodology to the challenging problems of autonomous driving and UAV\nracing. For both tasks, we utilize the Sim4CV simulator that enables the\ngeneration of large amounts of synthetic training data and also allows for\nonline learning and evaluation. We train a perception network to predict\nwaypoints from raw image data and use OIL to train another network to predict\ncontrols from these waypoints. Extensive experiments demonstrate that our\ntrained network outperforms its teachers, conventional imitation learning (IL)\nand reinforcement learning (RL) baselines and even humans in simulation. The\nproject website is available at this https URL\n",
"title": "OIL: Observational Imitation Learning"
} | null | null | null | null | true | null | 19826 | null | Default | null | null |
null | {
"abstract": " State space models in which the system state is a finite set--called the\nmulti-object state--have generated considerable interest in recent years.\nSmoothing for state space models provides better estimation performance than\nfiltering by using the full posterior rather than the filtering density. In\nmulti-object state estimation, the Bayes multi-object filtering recursion\nadmits an analytic solution known as the Generalized Labeled Multi-Bernoulli\n(GLMB) filter. In this work, we extend the analytic GLMB recursion to propagate\nthe multi-object posterior. We also propose an implementation of this so-called\nmulti-scan GLMB posterior recursion using a similar approach to the GLMB filter\nimplementation.\n",
"title": "A Multi-Scan Labeled Random Finite Set Model for Multi-object State Estimation"
} | null | null | [
"Statistics"
]
| null | true | null | 19827 | null | Validated | null | null |
null | {
"abstract": " We study a number of graph exploration problems in the following natural\nscenario: an algorithm starts exploring an undirected graph from some seed\nnode; the algorithm, for an arbitrary node $v$ that it is aware of, can ask an\noracle to return the set of the neighbors of $v$. (In social network analysis,\na call to this oracle corresponds to downloading the profile page of user $v$\nin a social network.) The goal of the algorithm is to either learn something\n(e.g., average degree) about the graph, or to return some random function of\nthe graph (e.g., a uniform-at-random node), while accessing/downloading as few\nnodes of the graph as possible. Motivated by practical applications, we study\nthe complexities of a variety of problems in terms of the graph's mixing time\nand average degree -- two measures that are believed to be quite small in\nreal-world social networks, and that have often been used in the applied\nliterature to bound the performance of online exploration algorithms. Our main\nresult is that the algorithm has to access $\\Omega\\left(t_{\\rm mix} d_{\\rm avg}\n\\epsilon^{-2} \\ln \\delta^{-1}\\right)$ nodes to obtain, with probability at\nleast $1-\\delta$, an $\\epsilon$-additive approximation of the average of a\nbounded function on the nodes of a graph -- this lower bound matches the\nperformance of an algorithm that was proposed in the literature. We also give\ntight bounds for the problem of returning a close-to-uniform-at-random node\nfrom the graph. Finally, we give lower bounds for the problems of estimating\nthe average degree of the graph, and the number of nodes of the graph.\n",
"title": "On the Complexity of Sampling Nodes Uniformly from a Graph"
} | null | null | null | null | true | null | 19828 | null | Default | null | null |
null | {
"abstract": " The Fourier-Walsh expansion of a Boolean function $f \\colon \\{0,1\\}^n\n\\rightarrow \\{0,1\\}$ is its unique representation as a multilinear polynomial.\nThe Kindler-Safra theorem (2002) asserts that if in the expansion of $f$, the\ntotal weight on coefficients beyond degree $k$ is very small, then $f$ can be\napproximated by a Boolean-valued function depending on at most $O(2^k)$\nvariables.\nIn this paper we prove a similar theorem for Boolean functions whose domain\nis the `slice' ${{[n]}\\choose{pn}} = \\{x \\in \\{0,1\\}^n\\colon \\sum_i x_i =\npn\\}$, where $0 \\ll p \\ll 1$, with respect to their unique representation as\nharmonic multilinear polynomials. We show that if in the representation of\n$f\\colon {{[n]}\\choose{pn}} \\rightarrow \\{0,1\\}$, the total weight beyond\ndegree $k$ is at most $\\epsilon$, where $\\epsilon = \\min(p, 1-p)^{O(k)}$, then\n$f$ can be $O(\\epsilon)$-approximated by a degree-$k$ Boolean function on the\nslice, which in turn depends on $O(2^{k})$ coordinates. This proves a\nconjecture of Filmus, Kindler, Mossel, and Wimmer (2015). Our proof relies on\nhypercontractivity, along with a novel kind of a shifting procedure.\nIn addition, we show that the approximation rate in the Kindler-Safra theorem\ncan be improved from $\\epsilon + \\exp(O(k)) \\epsilon^{1/4}$ to\n$\\epsilon+\\epsilon^2 (2\\ln(1/\\epsilon))^k/k!$, which is tight in terms of the\ndependence on $\\epsilon$ and misses at most a factor of $2^{O(k)}$ in the\nlower-order term.\n",
"title": "A structure theorem for almost low-degree functions on the slice"
} | null | null | null | null | true | null | 19829 | null | Default | null | null |
null | {
"abstract": " Oxygen-deficient TiO$_2$ in the rutile structure as well as the Ti$_3$O$_5$\nMagn{é}li phase is investigated within the charge self-consistent combination\nof density functional theory (DFT) with dynamical mean-field theory (DMFT). It\nis shown that an isolated oxygen vacancy (V$_{\\rm O}$) in titanium dioxide is\nnot sufficient to metallize the system at low temperatures. In a semiconducting\nphase, an in-gap state is identified at $\\varepsilon_{\\rm IG}^{\\hfill}\\sim\n-0.75\\,$eV\\, in excellent agreement with experimental data. Band-like impurity\nlevels, resulting from a threefold V$_{\\rm O}$-Ti coordination as well as\nentangled $(t_{2g},e_g)$ states, become localized due to site-dependent\nelectronic correlations. Charge localization and strong orbital polarization\noccur in the V$_{\\rm O}$-near Ti ions, which details can be modified by a\nvariation of the correlated subspace. At higher oxygen vacancy concentration, a\ncorrelated metal is stabilized in the Magn{é}li phase. A V$_{\\rm O}$-defect\nrutile structure of identical stoichiometry shows key differences in the\norbital-resolved character and the spectral properties. Charge\ndisproportionation is vital in the oxygen-deficient compounds, but obvious\nmetal-insulator transitions driven or sustained by charge order are not\nidentified.\n",
"title": "Oxygen-vacancy driven electron localization and itinerancy in rutile-based TiO$_2$"
} | null | null | null | null | true | null | 19830 | null | Default | null | null |
null | {
"abstract": " Auxetic materials are a novel class of mechanical metamaterials which exhibit\nan interesting property of negative Poisson ratio by virtue of their\narchitecture rather than composition. It has been well established that a wide\nrange of negative Poisson ratio can be obtained by varying the geometry and\narchitecture of the cellular materials. However, the limited range of stiffness\nvalues obtained from a given geometry restricts their applications. Research\ntrials have revealed that multi-material cellular designs have the capability\nto generate range of stiffness values as per the requirement of application.\nWith the advancements in 3D printing, multi-material cellular designs can be\nrealized in practice. In this work, multi-material cellular designs are\ninvestigated using finite element method. It was observed that introduction of\nmaterial gradient/distribution in the cell provides a means to tune cellular\nstiffness as per the specific requirement. These results will aid in the design\nof wearable auxetic impact protection devices which rely on stiffness gradients\nand variable auxeticity.\n",
"title": "3D printable multimaterial cellular auxetics with tunable stiffness"
} | null | null | null | null | true | null | 19831 | null | Default | null | null |
null | {
"abstract": " Exploiting the powerful tool of strong gravitational lensing by galaxy\nclusters to study the highest-redshift Universe and cluster mass distributions\nrelies on precise lens mass modelling. In this work, we present the first\nattempt at modelling line-of-sight mass distribution in addition to that of the\ncluster, extending previous modelling techniques that assume mass distributions\nto be on a single lens plane. We focus on the Hubble Frontier Field cluster\nMACS J0416.1-2403, and our multi-plane model reproduces the observed image\npositions with a rms offset of ~0.53\". Starting from this best-fitting model,\nwe simulate a mock cluster that resembles MACS J0416.1-2403 in order to explore\nthe effects of line-of-sight structures on cluster mass modelling. By\nsystematically analysing the mock cluster under different model assumptions, we\nfind that neglecting the lensing environment has a significant impact on the\nreconstruction of image positions (rms ~0.3\"); accounting for line-of-sight\ngalaxies as if they were at the cluster redshift can partially reduce this\noffset. Moreover, foreground galaxies are more important to include into the\nmodel than the background ones. While the magnification factors of the lensed\nmultiple images are recovered within ~10% for ~95% of them, those ~5% that lie\nnear critical curves can be significantly affected by the exclusion of the\nlensing environment in the models (up to a factor of ~200). In addition,\nline-of-sight galaxies cannot explain the apparent discrepancy in the\nproperties of massive subhalos between MACS J0416.1-2403 and N-body simulated\nclusters. Since our model of MACS J0416.1-2403 with line-of-sight galaxies only\nreduced modestly the rms offset in the image positions, we conclude that\nadditional complexities, such as more flexible halo shapes, would be needed in\nfuture models of MACS J0416.1-2403.\n",
"title": "MACS J0416.1-2403: Impact of line-of-sight structures on strong gravitational lensing modelling of galaxy clusters"
} | null | null | null | null | true | null | 19832 | null | Default | null | null |
null | {
"abstract": " Autonomous agents must often detect affordances: the set of behaviors enabled\nby a situation. Affordance detection is particularly helpful in domains with\nlarge action spaces, allowing the agent to prune its search space by avoiding\nfutile behaviors. This paper presents a method for affordance extraction via\nword embeddings trained on a Wikipedia corpus. The resulting word vectors are\ntreated as a common knowledge database which can be queried using linear\nalgebra. We apply this method to a reinforcement learning agent in a text-only\nenvironment and show that affordance-based action selection improves\nperformance most of the time. Our method increases the computational complexity\nof each learning step but significantly reduces the total number of steps\nneeded. In addition, the agent's action selections begin to resemble those a\nhuman would choose.\n",
"title": "What can you do with a rock? Affordance extraction via word embeddings"
} | null | null | null | null | true | null | 19833 | null | Default | null | null |
null | {
"abstract": " We revisit the question of whether and how the steady states arising after\nnon-equilibrium time evolution in integrable models (and in particular in the\nXXZ spin chain) can be described by the so-called Generalized Gibbs Ensemble\n(GGE). It is known that the micro-canonical ensemble built on a complete set of\ncharges correctly describes the long-time limit of local observables, and\nrecently a canonical ensemble was built by Ilievski et. al. using particle\noccupation number operators. Here we provide an alternative construction by\nconsidering truncated GGE's (tGGE's) that only include a finite number of well\nlocalized conserved operators. It is shown that the tGGE's can approximate the\nsteady states with arbitrary precision, i.e. all physical observables are\nexactly reproduced in the infinite truncation limit. In addition, we show that\na complete canonical ensemble can in fact be built in terms of a new (discrete)\nset of charges built as linear combinations of the standard ones.\nOur general arguments are applied to concrete quench situations in the XXZ\nchain, where the initial states are simple two-site or four-site product\nstates. Depending on the quench we find that numerical results for the local\ncorrelators can be obtained with remarkable precision using truncated GGE's\nwith only 10-100 charges.\n",
"title": "On Generalized Gibbs Ensembles with an infinite set of conserved charges"
} | null | null | null | null | true | null | 19834 | null | Default | null | null |
null | {
"abstract": " Trapping molecular ions that have been sympathetically cooled with\nlaser-cooled atomic ions is a useful platform for exploring cold ion chemistry.\nWe designed and characterized a new experimental apparatus for probing chemical\nreaction dynamics between molecular cations and neutral radicals at\ntemperatures below 1 K. The ions are trapped in a linear quadrupole\nradio-frequency trap and sympathetically cooled by co-trapped, laser-cooled,\natomic ions. The ion trap is coupled to a time-of-flight mass spectrometer to\nreadily identify product ion species, as well as to accurately determine\ntrapped ion numbers. We discuss, and present in detail, the design of this ion\ntrap time-of-flight mass spectrometer, as well as the electronics required for\ndriving the trap and mass spectrometer. Furthermore, we measure the performance\nof this system, which yields mass resolutions of $m/\\Delta{}m \\geq 1100$ over a\nwide mass range, and discuss its relevance for future measurements in chemical\nreaction kinetics and dynamics.\n",
"title": "High resolution ion trap time-of-flight mass spectrometer for cold trapped ion experiments"
} | null | null | null | null | true | null | 19835 | null | Default | null | null |
null | {
"abstract": " The precise nature of complex structural relaxation as well as an explanation\nfor the precipitous growth of relaxation time in cooling glass-forming liquids\nare essential to the understanding of vitrification of liquids. The dramatic\nincrease of relaxation time is believed to be caused by the growth of one or\nmore correlation lengths, which has received much attention recently. Here, we\nreport a direct link between the growth of a specific local-geometrical-order\nand an increase of dynamic-length-scale as the atomic dynamics in metallic\nglass-forming liquids slow down. Although several types of local\ngeometrical-orders are present in these metallic liquids, the growth of\nicosahedral ordering is found to be directly related to the increase of the\ndynamic-length-scale. This finding suggests an intriguing scenario that the\ntransient icosahedral ordering could be the origin of the dynamic-length-scale\nin metallic glass-forming liquids.\n",
"title": "The role of local-geometrical-orders on the growth of dynamic-length-scales in glass-forming liquids"
} | null | null | null | null | true | null | 19836 | null | Default | null | null |
null | {
"abstract": " A regularized optimization problem over a large unstructured graph is\nstudied, where the regularization term is tied to the graph geometry. Typical\nregularization examples include the total variation and the Laplacian\nregularizations over the graph. When applying the proximal gradient algorithm\nto solve this problem, there exist quite affordable methods to implement the\nproximity operator (backward step) in the special case where the graph is a\nsimple path without loops. In this paper, an algorithm, referred to as \"Snake\",\nis proposed to solve such regularized problems over general graphs, by taking\nbenefit of these fast methods. The algorithm consists in properly selecting\nrandom simple paths in the graph and performing the proximal gradient algorithm\nover these simple paths. This algorithm is an instance of a new general\nstochastic proximal gradient algorithm, whose convergence is proven.\nApplications to trend filtering and graph inpainting are provided among others.\nNumerical experiments are conducted over large graphs.\n",
"title": "Snake: a Stochastic Proximal Gradient Algorithm for Regularized Problems over Large Graphs"
} | null | null | null | null | true | null | 19837 | null | Default | null | null |
null | {
"abstract": " We analyze the effect of intersite-interaction terms on the stability of the\ncoexisting superconucting-nematic phase (SC+N) within the extended Hubbard and\n$t$-$J$-$U$ models on the square lattice. In order to take into account the\ncorrelation effects with a proper precision, we use the approach based on the\n\\textit{diagrammatic expansion of the Gutzwiller wave function} (DE-GWF), which\ngoes beyond the renormalized mean field theory (RMFT) in a systematic manner.\nAs a starting point of our analysis we discuss the stability region of the SC+N\nphase on the intrasite Coulomb repulsion-hole doping plane for the case of the\nHubbard model. Next, we show that the exchange interaction term enhances\nsuperconductivity while suppresses the nematicity, whereas the intersite\nCoulomb repulsion term acts in the opposite manner. The competing character of\nthe SC and N phases interplay is clearly visible throughout the analysis. A\nuniversal conclusion is that the nematic phase does not survive within the\n$t$-$J$-$U$ model with the value of $J$ integral typical for the high-T$_C$\ncuprates ($J\\approx 0.1$eV). For the sake of completeness, the effect of the\ncorrelated hopping term is also analyzed. Thus the present discussion contains\nall relevant two-site interaction terms which appear in the parametrized\none-band model within the second quantization scheme. At the end, the influence\nof the higher-order terms of the diagrammatic expansion on the rotational\nsymmetry breaking is also shown by comparing the DE-GWF results with those\ncorresponding to the RMFT.\n",
"title": "Stability of the coexistent superconducting-nematic phase under the presence of intersite interactions"
} | null | null | null | null | true | null | 19838 | null | Default | null | null |
null | {
"abstract": " We consider tackling a single-agent RL problem by distributing it to $n$\nlearners. These learners, called advisors, endeavour to solve the problem from\na different focus. Their advice, taking the form of action values, is then\ncommunicated to an aggregator, which is in control of the system. We show that\nthe local planning method for the advisors is critical and that none of the\nones found in the literature is flawless: the egocentric planning overestimates\nvalues of states where the other advisors disagree, and the agnostic planning\nis inefficient around danger zones. We introduce a novel approach called\nempathic and discuss its theoretical aspects. We empirically examine and\nvalidate our theoretical findings on a fruit collection task.\n",
"title": "Multi-Advisor Reinforcement Learning"
} | null | null | null | null | true | null | 19839 | null | Default | null | null |
null | {
"abstract": " The dissipation of small-scale perturbations in the early universe produces a\ndistortion in the blackbody spectrum of cosmic microwave background photons. In\nthis work, we propose to use these distortions as a probe of the microphysics\nof dark matter on scales $1\\,\\textrm{Mpc}^{-1}\\lesssim k \\lesssim\n10^{4}\\,\\textrm{Mpc}^{-1}$. We consider in particular models in which the dark\nmatter is kinetically coupled to either neutrinos or photons until shortly\nbefore recombination, and compute the photon heating rate and the resultant\n$\\mu$-distortion in both cases. We show that the $\\mu$-parameter is generally\nenhanced relative to $\\Lambda$CDM for interactions with neutrinos, and may be\neither enhanced or suppressed in the case of interactions with photons. The\ndeviations from the $\\Lambda$CDM signal are potentially within the sensitivity\nreach of a PRISM-like experiment if $\\sigma_{\\textrm{DM}-\\gamma} \\gtrsim\n1.1\\times10^{-30} \\left(m_{\\textrm{DM}}/\\textrm{GeV}\\right) \\textrm{cm}^{2}$\nand $\\sigma_{\\textrm{DM}-\\nu} \\gtrsim 4.8\\times 10^{-32}\n\\left(m_{\\textrm{DM}}/\\textrm{GeV}\\right) \\textrm{cm}^{2}$ for time-independent\ncross sections, and $\\sigma^{0}_{\\textrm{DM}-\\gamma} \\gtrsim 1.8 \\times\n10^{-40} \\left(m_{\\textrm{DM}}/\\textrm{GeV}\\right) \\textrm{cm}^{2}$ and\n$\\sigma^{0}_{\\textrm{DM}-\\nu} \\gtrsim 2.5 \\times 10^{-47}\n\\left(m_{\\textrm{DM}}/\\textrm{GeV}\\right) \\textrm{cm}^{2}$ for cross sections\nscaling as temperature squared, coinciding with the parameter regions in which\nlate kinetic decoupling may serve as a solution to the small-scale crisis.\nFurthermore, these $\\mu$-distortion signals differ from those of warm dark\nmatter (no deviation from $\\Lambda$CDM) and a suppressed primordial power\nspectrum (strongly suppressed or a negative $\\mu$-parameter), demonstrating\nthat CMB spectral distortion can potentially be used to distinguish between\nsolutions to the small-scale crisis.\n",
"title": "Using CMB spectral distortions to distinguish between dark matter solutions to the small-scale crisis"
} | null | null | null | null | true | null | 19840 | null | Default | null | null |
null | {
"abstract": " Multi-subject fMRI data analysis is an interesting and challenging problem in\nhuman brain decoding studies. The inherent anatomical and functional\nvariability across subjects make it necessary to do both anatomical and\nfunctional alignment before classification analysis. Besides, when it comes to\nbig data, time complexity becomes a problem that cannot be ignored. This paper\nproposes Gradient Hyperalignment (Gradient-HA) as a gradient-based functional\nalignment method that is suitable for multi-subject fMRI datasets with large\namounts of samples and voxels. The advantage of Gradient-HA is that it can\nsolve independence and high dimension problems by using Independent Component\nAnalysis (ICA) and Stochastic Gradient Ascent (SGA). Validation using\nmulti-classification tasks on big data demonstrates that Gradient-HA method has\nless time complexity and better or comparable performance compared with other\nstate-of-the-art functional alignment methods.\n",
"title": "Gradient Hyperalignment for multi-subject fMRI data alignment"
} | null | null | null | null | true | null | 19841 | null | Default | null | null |
null | {
"abstract": " Knowledge transfer between tasks can improve the performance of learned\nmodels, but requires an accurate estimate of the inter-task relationships to\nidentify the relevant knowledge to transfer. These inter-task relationships are\ntypically estimated based on training data for each task, which is inefficient\nin lifelong learning settings where the goal is to learn each consecutive task\nrapidly from as little data as possible. To reduce this burden, we develop a\nlifelong learning method based on coupled dictionary learning that utilizes\nhigh-level task descriptions to model the inter-task relationships. We show\nthat using task descriptors improves the performance of the learned task\npolicies, providing both theoretical justification for the benefit and\nempirical demonstration of the improvement across a variety of learning\nproblems. Given only the descriptor for a new task, the lifelong learner is\nalso able to accurately predict a model for the new task through zero-shot\nlearning using the coupled dictionary, eliminating the need to gather training\ndata before addressing the task.\n",
"title": "Using Task Descriptions in Lifelong Machine Learning for Improved Performance and Zero-Shot Transfer"
} | null | null | null | null | true | null | 19842 | null | Default | null | null |
null | {
"abstract": " Because the grand unification theory of gauge theories of strong, weak and\nelectromagnetic interactions is based on principal bundle theory, and\ngravitational theory is based on the tangent vector bundle theory, so people\ncannot unify the these four basic interactions in principal bundle theory. This\nLetter discovers and gives giant unification theory of the grand unification\ntheory and gravitation theory, i.e., the giant unification theory of strong,\nweak, electromagnetic and gravitational interactions according to the general\nfiber bundle theory, symmetry, quantitative causal principle (QCP) and so on.\nConsequently, the research of this Letter is based on the rigorous scientific\nbases of mathematics and physics. The Lagrangians of the well-known fundamental\nphysics interactions are unifiedly deduced from QCP and satisfy the gauge\ninvariant principle of general gauge fields interacting with Fermion and/or\nboson fields. The geometry and physics meanings of gauge invariant property of\ndifferent physical systems are revealed, and it is discovered that all the\nLagrangians of the well-known fundamental physics interactions are composed of\nthe invariant quantities in corresponding spacetime structures. The\ndifficulties that fundamental physics interactions and Noether theorem are not\nable to be unifiedly given and investigated are overcome, the unified\ndescription and origin of the fundamental physics interactions and Noether\ntheorem are shown by QCP, their two-order general Euler-Lagrange Equations and\ncorresponding Noether conservation currents are derived in general curved\nspacetime. Therefore, using the new unification theory, a lot of research works\nabout different branches of physics etc can be renewedly done and expressed\nsimpler with clear quantitative causal physical meanings.\n",
"title": "Giant Unification Theory of the Grand Unification and Gravitation Theories"
} | null | null | null | null | true | null | 19843 | null | Default | null | null |
null | {
"abstract": " Rich-club ordering refers to the tendency of nodes with a high degree to be\nmore interconnected than expected. In this paper we consider the concept of\nrich-club ordering when generalized to structural measures that differ from the\nnode degree and to non-structural measures (i.e. to node metadata). The\ndifferences in considering rich-club ordering (RCO) with respect to both\nstructural and non-structural measures is then discussed in terms of employed\ncoefficients and of appropriate null models (link rewiring vs metadata\nreshuffling). Once a framework for the evaluation of generalized rich-club\nordering (GRCO) is defined, we investigate such a phenomenon in real networks\nprovided with node metadata. By considering different notions of node richness,\nwe compare structural and non-structural rich-club ordering, observing how\nexternal information about the network nodes is able to validate the presence\nof rich-clubs in networked systems.\n",
"title": "Generalized Rich-Club Ordering in Networks"
} | null | null | null | null | true | null | 19844 | null | Default | null | null |
null | {
"abstract": " In this work, we propose to integrate prediction algorithms to the scheduling\nof mode changes under the Earliest-Deadline-First and Fixed-priority scheduling\nin mixed-criticality real-time systems. The method proactively schedules a mode\nchange in the system based on state variables such as laxity, to the percentage\ndifference in the temporal distance between the completion time of the instance\nof a task and its respective deadline, by the deadline (D) stipulated for the\ntask, in order to minimize deadline misses. The simulation model was validated\nagainst an analytical model prior to the logical integration of the\nKalman-based prediction algorithm. Two study cases were presented, one covering\nearliest-deadline first and the other the fixed-priority scheduling approach.\nThe results showed the gains in the adoption of the prediction approach for\nboth scheduling paradigms by presenting a significant reduction of the number\nof missed deadlines for low-criticality tasks.\n",
"title": "Integrating Proactive Mode Changes in Mixed Criticality Systems"
} | null | null | null | null | true | null | 19845 | null | Default | null | null |
null | {
"abstract": " Inverse electromagnetic design has emerged as a way of efficiently designing\nactive and passive electromagnetic devices. This maturing strategy involves\noptimizing the shape or topology of a device in order to improve a figure of\nmerit--a process which is typically performed using some form of steepest\ndescent algorithm. Naturally, this requires that we compute the gradient of a\nfigure of merit which describes device performance, potentially with respect to\nmany design variables. In this paper, we introduce a new strategy based on\nsmoothing abrupt material interfaces which enables us to efficiently compute\nthese gradients with high accuracy irrespective of the resolution of the\nunderlying simulation. This has advantages over previous approaches to shape\nand topology optimization in nanophotonics which are either prone to gradient\nerrors or place important constraints on the shape of the device. As a\ndemonstration of this new strategy, we optimize a non-adiabatic waveguide taper\nbetween a narrow and wide waveguide. This optimization leads to a non-intuitive\ndesign with a very low insertion loss of only 0.041 dB at 1550 nm.\n",
"title": "Leveraging Continuous Material Averaging for Inverse Electromagnetic Design"
} | null | null | null | null | true | null | 19846 | null | Default | null | null |
null | {
"abstract": " This paper describes a general, scalable, end-to-end framework that uses the\ngenerative adversarial network (GAN) objective to enable robust speech\nrecognition. Encoders trained with the proposed approach enjoy improved\ninvariance by learning to map noisy audio to the same embedding space as that\nof clean audio. Unlike previous methods, the new framework does not rely on\ndomain expertise or simplifying assumptions as are often needed in signal\nprocessing, and directly encourages robustness in a data-driven way. We show\nthe new approach improves simulated far-field speech recognition of vanilla\nsequence-to-sequence models without specialized front-ends or preprocessing.\n",
"title": "Robust Speech Recognition Using Generative Adversarial Networks"
} | null | null | null | null | true | null | 19847 | null | Default | null | null |
null | {
"abstract": " We show that maximal $S$-free convex sets are polyhedra when $S$ is the set\nof integral points in some rational polyhedron of $\\mathbb{R}^n$. This result\nextends a theorem of Lovász characterizing maximal lattice-free convex sets.\nOur theorem has implications in integer programming. In particular, we show\nthat maximal $S$-free convex sets are in one-to-one correspondence with minimal\ninequalities.\n",
"title": "Minimal inequalities for an infinite relaxation of integer programs"
} | null | null | null | null | true | null | 19848 | null | Default | null | null |
null | {
"abstract": " The nearby ultra-luminous infrared galaxy (ULIRG) Arp 220 is an excellent\nlaboratory for studies of extreme astrophysical environments. For 20 years,\nVery Long Baseline Interferometry (VLBI) has been used to monitor a population\nof compact sources thought to be supernovae (SNe), supernova remnants (SNRs)\nand possibly active galactic nuclei (AGNs). Using new and archival VLBI data\nspanning 20 years, we obtain 23 high-resolution radio images of Arp 220 at\nwavelengths from 18 cm to 2 cm. From model-fitting to the images we obtain\nestimates of flux densities and sizes of all detected sources. We detect radio\ncontinuum emission from 97 compact sources and present flux densities and sizes\nfor all analysed observation epochs. We find evidence for a LD-relation within\nArp 220, with larger sources being less luminous. We find a compact source LF\n$n(L)\\propto L^\\beta$ with $\\beta=-2.02\\pm0.11$, similar to SNRs in normal\ngalaxies. Based on simulations we argue that there are many relatively large\nand weak sources below our detection threshold. The rapidly declining object\n0.2227+0.482 is proposed as a possible AGN candidate. The observations can be\nexplained by a mixed population of SNe and SNRs, where the former expand in a\ndense circumstellar medium (CSM) and the latter interact with the surrounding\ninterstellar medium (ISM). Several sources (9 of the 94 with fitted sizes) are\nlikely luminous, type IIn SNe. This number of luminous SNe correspond to few\npercent of the total number of SNe in Arp 220 which is consistent with a total\nSN-rate of 4 yr$^{-1}$ as inferred from the total radio emission given a normal\nstellar initial mass function (IMF). Based on the fitted luminosity function,\nwe argue that emission from all compact sources, also below our detection\nthreshold, make up at most 20\\% of the total radio emission at GHz frequencies.\n",
"title": "The population of SNe/SNRs in the starburst galaxy Arp 220. A self-consistent analysis of 20 years of VLBI monitoring"
} | null | null | [
"Physics"
]
| null | true | null | 19849 | null | Validated | null | null |
null | {
"abstract": " Motivated by recent experimental observations in $\\alpha$-RuCl$_3$, we study\nthe $K$-$\\Gamma$ model on the honeycomb lattice in an external magnetic field.\nBy a slave-particle representation and Variational Monte Carlo calculations, we\nreproduce the phase transition from zigzag magnetic order to a field-induced\ndisordered phase. The nature of this state depends crucially on the field\norientation. For particular field directions in the honeycomb plane, we find a\ngapless Dirac spin liquid, in agreement with recent experiments on\n$\\alpha$-RuCl$_3$. For a range of out-of-plane fields, we predict the existence\nof a Kalmeyer-Laughlin-type chiral spin liquid, which would show an\ninteger-quantized thermal Hall effect.\n",
"title": "Dirac and Chiral Quantum Spin Liquids on the Honeycomb Lattice in a Magnetic Field"
} | null | null | [
"Physics"
]
| null | true | null | 19850 | null | Validated | null | null |
null | {
"abstract": " The behavior of an interior test particle in the secular 3-body problem has\nbeen studied extensively. A well-known feature is the Lidov-Kozai resonance in\nwhich the test particle's argument of periapse librates about $\\pm 90^\\circ$\nand large oscillations in eccentricity and inclination are possible. Less\nexplored is the inverse problem: the dynamics of an exterior test particle and\nan interior perturber. We survey numerically the inverse secular problem,\nexpanding the potential to hexadecapolar order and correcting an error in the\npublished expansion. Four secular resonances are uncovered that persist in full\n$N$-body treatments (in what follows, $\\varpi$ and $\\Omega$ are the longitudes\nof periapse and of ascending node, $\\omega$ is the argument of periapse, and\nsubscripts 1 and 2 refer to the inner perturber and outer test particle): (i)\nan orbit-flipping quadrupole resonance requiring a non-zero perturber\neccentricity $e_1$, in which $\\Omega_2-\\varpi_1$ librates about $\\pm 90^\\circ$;\n(ii) a hexadecapolar resonance (the \"inverse Kozai\" resonance) for perturbers\nthat are circular or nearly so and inclined by $I \\simeq 63^\\circ/117^\\circ$,\nin which $\\omega_2$ librates about $\\pm 90^\\circ$ and which can vary the\nparticle eccentricity by $\\Delta e_2 \\simeq 0.2$ and lead to orbit crossing;\n(iii) an octopole \"apse-aligned\" resonance at $I \\simeq 46^\\circ/107^\\circ$\nwherein $\\varpi_2 - \\varpi_1$ librates about $0^\\circ$ and $\\Delta e_2$ grows\nwith $e_1$; and (iv) an octopole resonance at $I \\simeq 73^\\circ/134^\\circ$\nwherein $\\varpi_2 + \\varpi_1 - 2 \\Omega_2$ librates about $0^\\circ$ and $\\Delta\ne_2$ can be as large as 0.3 for small $e_1 \\neq 0$. The more eccentric the\nperturber, the more the particle's eccentricity and inclination vary; also,\nmore polar orbits are more chaotic. Our inverse solutions may be applied to the\nKuiper belt and debris disks, circumbinary planets, and stellar systems.\n",
"title": "Secular Dynamics of an Exterior Test Particle: The Inverse Kozai and Other Eccentricity-Inclination Resonances"
} | null | null | null | null | true | null | 19851 | null | Default | null | null |
null | {
"abstract": " The KATRIN experiment aims to determine the neutrino mass scale with a\nsensitivity of 200 meV/c^2 (90% C.L.) by a precision measurement of the shape\nof the tritium $\\beta$-spectrum in the endpoint region. The energy analysis of\nthe decay electrons is achieved by a MAC-E filter spectrometer. To determine\nthe transmission properties of the KATRIN main spectrometer, a mono-energetic\nand angular-selective electron source has been developed. In preparation for\nthe second commissioning phase of the main spectrometer, a measurement phase\nwas carried out at the KATRIN monitor spectrometer where the device was\noperated in a MAC-E filter setup for testing. The results of these measurements\nare compared with simulations using the particle-tracking software\n\"Kassiopeia\", which was developed in the KATRIN collaboration over recent\nyears.\n",
"title": "A pulsed, mono-energetic and angular-selective UV photo-electron source for the commissioning of the KATRIN experiment"
} | null | null | null | null | true | null | 19852 | null | Default | null | null |
null | {
"abstract": " The stabilization of lasers to absolute frequency references is a fundamental\nrequirement in several areas of atomic, molecular and optical physics. A range\nof techniques are available to produce a suitable reference onto which one can\n'lock' the laser, many of which depend on the specific internal structure of\nthe reference or are sensitive to laser intensity noise. We present a novel\nmethod using the frequency modulation of an acousto-optic modulator's carrier\n(drive) signal to generate two spatially separated beams, with a frequency\ndifference of only a few MHz. These beams are used to probe a narrow absorption\nfeature and the difference in their detected signals leads to a dispersion-like\nfeature suitable for wavelength stabilization of a diode laser. This simple and\nversatile method only requires a narrow absorption line and is therefore\nsuitable for both atomic and cavity based stabilization schemes. To demonstrate\nthe suitability of this method we lock an external cavity diode laser near the\n$^{85}\\mathrm{Rb}\\,5S_{1/2}\\rightarrow5P_{3/2}, F=3\\rightarrow F^{\\prime}=4$\nusing sub-Doppler pump probe spectroscopy and also demonstrate excellent\nagreement between the measured signal and a theoretical model.\n",
"title": "Carrier frequency modulation of an acousto-optic modulator for laser stabilization"
} | null | null | null | null | true | null | 19853 | null | Default | null | null |
null | {
"abstract": " In this paper, a new type of 3D bin packing problem (BPP) is proposed, in\nwhich a number of cuboid-shaped items must be put into a bin one by one\northogonally. The objective is to find a way to place these items that can\nminimize the surface area of the bin. This problem is based on the fact that\nthere is no fixed-sized bin in many real business scenarios and the cost of a\nbin is proportional to its surface area. Our research shows that this problem\nis NP-hard. Based on previous research on 3D BPP, the surface area is\ndetermined by the sequence, spatial locations and orientations of items. Among\nthese factors, the sequence of items plays a key role in minimizing the surface\narea. Inspired by recent achievements of deep reinforcement learning (DRL)\ntechniques, especially Pointer Network, on combinatorial optimization problems\nsuch as TSP, a DRL-based method is applied to optimize the sequence of items to\nbe packed into the bin. Numerical results show that the method proposed in this\npaper achieve about 5% improvement than heuristic method.\n",
"title": "Solving a New 3D Bin Packing Problem with Deep Reinforcement Learning Method"
} | null | null | [
"Computer Science"
]
| null | true | null | 19854 | null | Validated | null | null |
null | {
"abstract": " The goal of this article is to clarify the meaning of Computational Thinking.\nWe differentiate logical from computational reasoning and discuss the\nimportance of Computational Thinking in solving problems. The three pillars of\nComputational Thinking - Abstraction, Automation and Analysis - are outlined,\nhighlighting the role of each one in developing the skills needed for the\nproblem-solving process.\n-----\nO objetivo deste artigo é esclarecer o significado de Pensamento\nComputacional. Diferencia-se o raciocínio lógico do computacional e\ndiscute-se a importância do Pensamento Computacional na resolução de\nproblemas. Os três pilares do Pensamento Computacional - Abstração,\nAutomação e Análise - são delineados, destacando-se o papel de cada\num deles no desenvolvimento das habilidades necessárias para o processo de\nsolução de problemas.\n",
"title": "Entendendo o Pensamento Computacional"
} | null | null | null | null | true | null | 19855 | null | Default | null | null |
null | {
"abstract": " The problem of the estimation of relevance to a set of histograms generated\nby samples of a discrete time process is discussed on the base of the\nvariational principles proposed in the previous paper [1]. Some conditions for\ndimension reduction of corresponding linear programming problems are presented\nalso.\n",
"title": "Estimation of samples relevance by their histograms"
} | null | null | null | null | true | null | 19856 | null | Default | null | null |
null | {
"abstract": " In this paper we present a strategy for the the synthesis of acoustic sources\nwith controllable near fields in free space and finite depth homogeneous ocean\nenvironments. We first present the theoretical results at the basis of our\ndiscussion and then, to illustrate our findings we focus on the following three\nparticular examples:\n1. acoustic source approximating a prescribed field pattern in a given\nbounded sub- region of its near field. 2. acoustic source approximating\ndifferent prescribed field patterns in given disjoint bounded near field\nsub-regions. 3. acoustic source approximating a prescribed back-propagating\nfield in a given bounded near field sub-region while maintaining a very low far\nfield signature.\nFor each of these three examples, we discuss the optimization scheme used to\napprox- imate their solutions and support our claims through relevant numerical\nsimulations.\n",
"title": "On the synthesis of acoustic sources with controllable near fields"
} | null | null | null | null | true | null | 19857 | null | Default | null | null |
null | {
"abstract": " This paper presents a method of reconstruction a primary structure of a\nprotein that folds into a given geometrical shape. This method predicts the\nprimary structure of a protein and restores its linear sequence of amino acids\nin the polypeptide chain using the tertiary structure of a molecule. Unknown\namino acids are determined according to the principle of energy minimization.\nThis study represents inverse folding problem as a quadratic optimization\nproblem and uses different relaxation techniques to reduce it to the problem of\nconvex optimizations. Computational experiment compares the quality of these\napproaches on real protein structures.\n",
"title": "Inverse Protein Folding Problem via Quadratic Programming"
} | null | null | null | null | true | null | 19858 | null | Default | null | null |
null | {
"abstract": " We consider the problem of performing Spoken Language Understanding (SLU) on\nsmall devices typical of IoT applications. Our contributions are twofold.\nFirst, we outline the design of an embedded, private-by-design SLU system and\nshow that it has performance on par with cloud-based commercial solutions.\nSecond, we release the datasets used in our experiments in the interest of\nreproducibility and in the hope that they can prove useful to the SLU\ncommunity.\n",
"title": "Spoken Language Understanding on the Edge"
} | null | null | [
"Computer Science"
]
| null | true | null | 19859 | null | Validated | null | null |
null | {
"abstract": " We present a method to derive the density scaling relation $\\langle n\\rangle\n\\propto L^{-\\alpha}$ in regions of star formation or in their turbulent\nvicinities from straightforward binning of the column-density distribution\n($N$-pdf). The outcome of the method is studied for three types of $N$-pdf:\npower law ($7/5\\le\\alpha\\le5/3$), lognormal ($0.7\\lesssim\\alpha\\lesssim1.4$)\nand combination of lognormals. In the last case, the method of Stanchev et al.\n(2015) was also applied for comparison and a very weak (or close to zero)\ncorrelation was found. We conclude that the considered `binning approach'\nreflects rather the local morphology of the $N$-pdf with no reference to the\nphysical conditions in a considered region. The rough consistency of the\nderived slopes with the widely adopted Larson's (1981) value $\\alpha\\sim1.1$ is\nsuggested to support claims that the density-size relation in molecular clouds\nis indeed an artifact of the observed $N$-pdf.\n",
"title": "On the link between column density distribution and density scaling relation in star formation regions"
} | null | null | [
"Physics"
]
| null | true | null | 19860 | null | Validated | null | null |
null | {
"abstract": " Advertisements are unavoidable in modern society. Times Square is notorious\nfor its incessant display of advertisements. Its popularity is worldwide and\nsmaller cities possess miniature versions of the display, such as Pittsburgh\nand its digital works in Oakland on Forbes Avenue. Tokyo's Ginza district\nrecently rose to popularity due to its upscale shops and constant onslaught of\nadvertisements to pedestrians. Advertisements arise in other mediums as well.\nFor example, they help popular streaming services, such as Spotify, Hulu, and\nYoutube TV gather significant streams of revenue to reduce the cost of monthly\nsubscriptions for consumers. Ads provide an additional source of money for\ncompanies and entire industries to allocate resources toward alternative\nbusiness motives. They are attractive to companies and nearly unavoidable for\nconsumers. One challenge for advertisers is examining a advertisement's\neffectiveness or usefulness in conveying a message to their targeted\ndemographics. Rather than constructing a single, static image of content, a\nvideo advertisement possesses hundreds of frames of data with varying scenes,\nactors, objects, and complexity. Therefore, measuring effectiveness of video\nadvertisements is important to impacting a billion-dollar industry. This paper\nexplores the combination of human-annotated features and common video\nprocessing techniques to predict effectiveness ratings of advertisements\ncollected from Youtube. This task is seen as a binary (effective vs.\nnon-effective), four-way, and five-way machine learning classification task.\nThe first findings in terms of accuracy and inference on this dataset, as well\nas some of the first ad research, on a small dataset are presented. Accuracies\nof 84\\%, 65\\%, and 55\\% are reached on the binary, four-way, and five-way tasks\nrespectively.\n",
"title": "Measuring Effectiveness of Video Advertisements"
} | null | null | [
"Computer Science"
]
| null | true | null | 19861 | null | Validated | null | null |
null | {
"abstract": " We study a deep linear network endowed with a structure. It takes the form of\na matrix $X$ obtained by multiplying $K$ matrices (called factors and\ncorresponding to the action of the layers). The action of each layer (i.e. a\nfactor) is obtained by applying a fixed linear operator to a vector of\nparameters satisfying a constraint. The number of layers is not limited.\nAssuming that $X$ is given and factors have been estimated, the error between\nthe product of the estimated factors and $X$ (i.e. the reconstruction error) is\neither the statistical or the empirical risk. In this paper, we provide\nnecessary and sufficient conditions on the network topology under which a\nstability property holds. The stability property requires that the error on the\nparameters defining the factors (i.e. the stability of the recovered\nparameters) scales linearly with the reconstruction error (i.e. the risk).\nTherefore, under these conditions on the network topology, any successful\nlearning task leads to stably defined features and therefore interpretable\nlayers/network.In order to do so, we first evaluate how the Segre embedding and\nits inverse distort distances. Then, we show that any deep structured linear\nnetwork can be cast as a generic multilinear problem (that uses the Segre\nembedding). This is the {\\em tensorial lifting}. Using the tensorial lifting,\nwe provide necessary and sufficient conditions for the identifiability of the\nfactors (up to a scale rearrangement). We finally provide the necessary and\nsufficient condition called \\NSPlong~(because of the analogy with the usual\nNull Space Property in the compressed sensing framework) which guarantees that\nthe stability property holds. We illustrate the theory with a practical example\nwhere the deep structured linear network is a convolutional linear network. As\nexpected, the conditions are rather strong but not empty. A simple test on the\nnetwork topology can be implemented to test if the condition holds.\n",
"title": "Multilinear compressive sensing and an application to convolutional linear networks"
} | null | null | null | null | true | null | 19862 | null | Default | null | null |
null | {
"abstract": " We apply our theory of partial flag spaces developed in previous articles to\nstudy a group-theoretical generalization of the canonical filtration of a\ntruncated Barsotti-Tate group of level 1. As an application, we determine\nexplicitly the normalization of the Zariski closures of Ekedahl-Oort strata of\nShimura varieties of Hodge-type as certain closed coarse strata in the\nassociated partial flag spaces.\n",
"title": "Normalization of closed Ekedahl-Oort strata"
} | null | null | null | null | true | null | 19863 | null | Default | null | null |
null | {
"abstract": " The robustness of dynamical properties of neuronal networks against\nstructural damages is a central problem in computational and experimental\nneuroscience. Research has shown that the cortical network of a healthy brain\nworks near a critical state, and moreover, that functional neuronal networks\noften have scale-free and small-world properties. In this work, we study how\nthe robustness of simple functional networks at criticality is affected by\nstructural defects. In particular, we consider a 2D Ising model at the critical\ntemperature and investigate how its functional network changes with the\nincreasing degree of structural defects. We show that the scale-free and\nsmall-world properties of the functional network at criticality are robust\nagainst large degrees of structural lesions while the system remains below the\npercolation limit. Although the Ising model is only a conceptual description of\na two-state neuron, our research reveals fundamental robustness properties of\nfunctional networks derived from classical statistical mechanics models.\n",
"title": "Robustness of functional networks at criticality against structural defects"
} | null | null | null | null | true | null | 19864 | null | Default | null | null |
null | {
"abstract": " Maximizing the area under the receiver operating characteristic curve (AUC)\nis a standard approach to imbalanced classification. So far, various supervised\nAUC optimization methods have been developed and they are also extended to\nsemi-supervised scenarios to cope with small sample problems. However, existing\nsemi-supervised AUC optimization methods rely on strong distributional\nassumptions, which are rarely satisfied in real-world problems. In this paper,\nwe propose a novel semi-supervised AUC optimization method that does not\nrequire such restrictive assumptions. We first develop an AUC optimization\nmethod based only on positive and unlabeled data (PU-AUC) and then extend it to\nsemi-supervised learning by combining it with a supervised AUC optimization\nmethod. We theoretically prove that, without the restrictive distributional\nassumptions, unlabeled data contribute to improving the generalization\nperformance in PU and semi-supervised AUC optimization methods. Finally, we\ndemonstrate the practical usefulness of the proposed methods through\nexperiments.\n",
"title": "Semi-Supervised AUC Optimization based on Positive-Unlabeled Learning"
} | null | null | null | null | true | null | 19865 | null | Default | null | null |
null | {
"abstract": " Motivated by problems in data clustering, we establish general conditions\nunder which families of nonparametric mixture models are identifiable, by\nintroducing a novel framework involving clustering overfitted \\emph{parametric}\n(i.e. misspecified) mixture models. These identifiability conditions generalize\nexisting conditions in the literature, and are flexible enough to include for\nexample mixtures of Gaussian mixtures. In contrast to the recent literature on\nestimating nonparametric mixtures, we allow for general nonparametric mixture\ncomponents, and instead impose regularity assumptions on the underlying mixing\nmeasure. As our primary application, we apply these results to partition-based\nclustering, generalizing the notion of a Bayes optimal partition from classical\nparametric model-based clustering to nonparametric settings. Furthermore, this\nframework is constructive so that it yields a practical algorithm for learning\nidentified mixtures, which is illustrated through several examples on real\ndata. The key conceptual device in the analysis is the convex, metric geometry\nof probability measures on metric spaces and its connection to the Wasserstein\nconvergence of mixing measures. The result is a flexible framework for\nnonparametric clustering with formal consistency guarantees.\n",
"title": "Identifiability of Nonparametric Mixture Models and Bayes Optimal Clustering"
} | null | null | null | null | true | null | 19866 | null | Default | null | null |
null | {
"abstract": " We describe all rigid algebras and all irreducible components in the variety\nof four dimensional Leibniz algebras $\\mathfrak{Leib}_4$ over $\\mathbb{C}.$ In\nparticular, we prove that the Grunewald--O'Halloran conjecture is not valid and\nthe Vergne conjecture is valid for $\\mathfrak{Leib}_4.$\n",
"title": "The geometric classification of Leibniz algebras"
} | null | null | [
"Mathematics"
]
| null | true | null | 19867 | null | Validated | null | null |
null | {
"abstract": " In this paper we compute the discrete fundamental groups of warped cones. As\nan immediate consequence, this allows us to show that there exist coarsely\nsimply-connected expanders and superexpanders. This also provides a strong\ncoarse invariant of warped cones and implies that many warped cones cannot be\ncoarsely equivalent to any box space.\n",
"title": "Discrete fundamental groups of Warped Cones and expanders"
} | null | null | null | null | true | null | 19868 | null | Default | null | null |
null | {
"abstract": " We investigate the theoretical foundations of the simulated tempering method\nand use our findings to design efficient algorithms. Employing a large\ndeviation argument first used for replica exchange molecular dynamics [Plattner\net al., J. Chem. Phys. 135:134111 (2011)], we demonstrate that the most\nefficient approach to simulated tempering is to vary the temperature infinitely\nrapidly. In this limit, we can replace the equations of motion for the\ntemperature and physical variables by averaged equations for the latter alone,\nwith the forces rescaled according to a position-dependent function defined in\nterms of temperature weights. The averaged equations are similar to those used\nin Gao's integrated-over-temperature method, except that we show that it is\nbetter to use a continuous rather than a discrete set of temperatures. We give\na theoretical argument for the choice of the temperature weights as the\nreciprocal partition function, thereby relating simulated tempering to\nWang-Landau sampling. Finally, we describe a self-consistent algorithm for\nsimultaneously sampling the canonical ensemble and learning the weights during\nsimulation. This algorithm is tested on a system of harmonic oscillators as\nwell as a continuous variant of the Curie-Weiss model, where it is shown to\nperform well and to accurately capture the second-order phase transition\nobserved in this model.\n",
"title": "Simulated Tempering Method in the Infinite Switch Limit with Adaptive Weight Learning"
} | null | null | null | null | true | null | 19869 | null | Default | null | null |
null | {
"abstract": " We live in a 3D world, performing activities and interacting with objects in\nthe indoor environments everyday. Indoor scenes are the most familiar and\nessential environments in everyone's life. In the virtual world, 3D indoor\nscenes are also ubiquitous in 3D games and interior design. With the fast\ndevelopment of VR/AR devices and the emerging applications, the demand of\nrealistic 3D indoor scenes keeps growing rapidly. Currently, designing detailed\n3D indoor scenes requires proficient 3D designing and modeling skills and is\noften time-consuming. For novice users, creating realistic and complex 3D\nindoor scenes is even more difficult and challenging.\nMany efforts have been made in different research communities, e.g. computer\ngraphics, vision and robotics, to capture, analyze and generate the 3D indoor\ndata. This report mainly focuses on the recent research progress in graphics on\ngeometry, structure and semantic analysis of 3D indoor data and different\nmodeling techniques for creating plausible and realistic indoor scenes. We\nfirst review works on understanding and semantic modeling of scenes from\ncaptured 3D data of the real world. Then, we focus on the virtual scenes\ncomposed of 3D CAD models and study methods for 3D scene analysis and\nprocessing. After that, we survey various modeling paradigms for creating 3D\nindoor scenes and investigate human-centric scene analysis and modeling, which\nbridge indoor scene studies of graphics, vision and robotics. At last, we\ndiscuss open problems in indoor scene processing that might bring interests to\ngraphics and all related communities.\n",
"title": "Analysis and Modeling of 3D Indoor Scenes"
} | null | null | null | null | true | null | 19870 | null | Default | null | null |
null | {
"abstract": " Advances in astronomy are intimately linked to advances in digital signal\nprocessing (DSP). This special issue is focused upon advances in DSP within\nradio astronomy. The trend within that community is to use off-the-shelf\ndigital hardware where possible and leverage advances in high performance\ncomputing. In particular, graphics processing units (GPUs) and field\nprogrammable gate arrays (FPGAs) are being used in place of\napplication-specific circuits (ASICs); high-speed Ethernet and Infiniband are\nbeing used for interconnect in place of custom backplanes. Further, to lower\nhurdles in digital engineering, communities have designed and released\ngeneral-purpose FPGA-based DSP systems, such as the CASPER ROACH board, ASTRON\nUniboard and CSIRO Redback board. In this introductory article, we give a brief\nhistorical overview, a summary of recent trends, and provide an outlook on\nfuture directions.\n",
"title": "Introduction to the Special Issue on Digital Signal Processing in Radio Astronomy"
} | null | null | [
"Physics"
]
| null | true | null | 19871 | null | Validated | null | null |
null | {
"abstract": " Eliminating duplicate data in primary storage of clouds increases the\ncost-efficiency of cloud service providers as well as reduces the cost of users\nfor using cloud services. Existing primary deduplication techniques either use\ninline caching to exploit locality in primary workloads or use post-processing\ndeduplication running in system idle time to avoid the negative impact on I/O\nperformance. However, neither of them works well in the cloud servers running\nmultiple services or applications for the following two reasons: Firstly, the\ntemporal locality of duplicate data writes may not exist in some primary\nstorage workloads thus inline caching often fails to achieve good deduplication\nratio. Secondly, the post-processing deduplication allows duplicate data to be\nwritten into disks, therefore does not provide the benefit of I/O deduplication\nand requires high peak storage capacity. This paper presents HPDedup, a Hybrid\nPrioritized data Deduplication mechanism to deal with the storage system shared\nby applications running in co-located virtual machines or containers by fusing\nan inline and a post-processing process for exact deduplication. In the inline\ndeduplication phase, HPDedup gives a fingerprint caching mechanism that\nestimates the temporal locality of duplicates in data streams from different\nVMs or applications and prioritizes the cache allocation for these streams\nbased on the estimation. HPDedup also allows different deduplication threshold\nfor streams based on their spatial locality to reduce the disk fragmentation.\nThe post-processing phase removes duplicates whose fingerprints are not able to\nbe cached due to the weak temporal locality from disks. Our experimental\nresults show that HPDedup clearly outperforms the state-of-the-art primary\nstorage deduplication techniques in terms of inline cache efficiency and\nprimary deduplication efficiency.\n",
"title": "HPDedup: A Hybrid Prioritized Data Deduplication Mechanism for Primary Storage in the Cloud"
} | null | null | null | null | true | null | 19872 | null | Default | null | null |
null | {
"abstract": " Freeplay is a significant source of nonlinearity in aeroelastic systems and\nis strictly regulated by airworthiness authorities. It splits the phase plane\nof such systems into three piecewise linear subdomains. Depending on the\nlocation of the freeplay, limit cycle oscillations can result that span either\ntwo or three of these subdomains. The purpose of this work is to demonstrate\nthe existence of two-domain cycles both theoretically and experimentally. A\nsimple aeroelastic system with pitch, plunge and control deflection degrees of\nfreedom is investigated in the presence of freeplay in pitch. It is shown that\ntwo-domain and three-domain cycles can result from a grazing bifurcation and\npropagate in the decreasing airspeed direction. Close to the bifurcation, the\ntwo limit cycle branches interact with each other and aperiodic oscillations\nensue. Equivalent linearization is used to derive the conditions of existence\nof each type of limit cycle and to predict their amplitudes and frequencies.\nComparisons with measurements from wind tunnel experiments demonstrate that the\ntheory describes these phenomena with accuracy.\n",
"title": "Two-domain and three-domain limit cycles in a typical aeroelastic system with freeplay in pitch"
} | null | null | [
"Physics"
]
| null | true | null | 19873 | null | Validated | null | null |
null | {
"abstract": " Understanding user preference is essential to the optimization of recommender\nsystems. As a feedback of user's taste, rating scores can directly reflect the\npreference of a given user to a given product. Uncovering the latent components\nof user ratings is thus of significant importance for learning user interests.\nIn this paper, a new recommendation approach, called LCR, was proposed by\ninvestigating the latent components of user ratings. The basic idea is to\ndecompose an existing rating into several components via a cost-sensitive\nlearning strategy. Specifically, each rating is assigned to several latent\nfactor models and each model is updated according to its predictive errors.\nAfterwards, these accumulated predictive errors of models are utilized to\ndecompose a rating into several components, each of which is treated as an\nindependent part to retrain the latent factor models. Finally, all latent\nfactor models are combined linearly to estimate predictive ratings for users.\nIn contrast to existing methods, LCR provides an intuitive preference modeling\nstrategy via multiple component analysis at an individual perspective.\nMeanwhile, it is verified by the experimental results on several benchmark\ndatasets that the proposed method is superior to the state-of-art methods in\nterms of recommendation accuracy.\n",
"title": "Preference Modeling by Exploiting Latent Components of Ratings"
} | null | null | [
"Computer Science"
]
| null | true | null | 19874 | null | Validated | null | null |
null | {
"abstract": " Low-dimensional embeddings of nodes in large graphs have proved extremely\nuseful in a variety of prediction tasks, from content recommendation to\nidentifying protein functions. However, most existing approaches require that\nall nodes in the graph are present during training of the embeddings; these\nprevious approaches are inherently transductive and do not naturally generalize\nto unseen nodes. Here we present GraphSAGE, a general, inductive framework that\nleverages node feature information (e.g., text attributes) to efficiently\ngenerate node embeddings for previously unseen data. Instead of training\nindividual embeddings for each node, we learn a function that generates\nembeddings by sampling and aggregating features from a node's local\nneighborhood. Our algorithm outperforms strong baselines on three inductive\nnode-classification benchmarks: we classify the category of unseen nodes in\nevolving information graphs based on citation and Reddit post data, and we show\nthat our algorithm generalizes to completely unseen graphs using a multi-graph\ndataset of protein-protein interactions.\n",
"title": "Inductive Representation Learning on Large Graphs"
} | null | null | null | null | true | null | 19875 | null | Default | null | null |
null | {
"abstract": " Assume that $f:(\\mathbb{C}^n,0) \\to (\\mathbb{C},0)$ is an analytic function\ngerm at the origin with only isolated singularity. Let $\\mu$ and $\\tau$ be the\ncorresponding Milnor and Tjurina numbers. We show that $\\dfrac{\\mu}{\\tau} \\leq\nn$. As an application, we give a lower bound for the Tjurina number in terms of\n$n$ and the multiplicity of $f$ at the origin.\n",
"title": "Milnor and Tjurina numbers for a hypersurface germ with isolated singularity"
} | null | null | null | null | true | null | 19876 | null | Default | null | null |
null | {
"abstract": " We study the convergence properties of the Gibbs Sampler in the context of\nposterior distributions arising from Bayesian analysis of Gaussian hierarchical\nmodels. We consider centred and non-centred parameterizations as well as their\nhybrids including the full family of partially non-centred parameterizations.\nWe develop a novel methodology based on multi-grid decompositions to derive\nanalytic expressions for the convergence rates of the algorithm for an\narbitrary number of layers in the hierarchy, while previous work was typically\nlimited to the two-level case. Our work gives a complete understanding for the\nthree-level symmetric case and this gives rise to approximations for the\nnon-symmetric case. We also give analogous, if less explicit, results for\nmodels of arbitrary level. This theory gives rise to simple and\neasy-to-implement guidelines for the practical implementation of Gibbs samplers\non conditionally Gaussian hierarchical models.\n",
"title": "Analysis of the Gibbs Sampler for Gaussian hierarchical models via multigrid decomposition"
} | null | null | null | null | true | null | 19877 | null | Default | null | null |
null | {
"abstract": " We point out a unique mechanism to produce the relic abundance for glueball\ndark matter from a gauged $SU(N)_d$ hidden sector which is bridged to the\nstandard model sector through heavy vectorlike quarks colored under gauge\ninteractions from both sides. A necessary ingredient of our assumption is that\nthe vectorlike quarks, produced either thermally or non-thermally, are abundant\nenough to dominate the universe for some time in the early universe. They later\nundergo dark color confinement and form unstable vectorlike-quarkonium states\nwhich annihilate decay and reheat the visible and dark sectors. The ratio of\nentropy dumped into two sectors and the final energy budget in the dark\nglueballs is only determined by low energy parameters, including the intrinsic\nscale of the dark $SU(N)_d$, $\\Lambda_d$, and number of dark colors, $N_d$, but\ndepend weakly on parameters in the ultraviolet such as the vectorlike quark\nmass or the initial condition. We call this a cosmic selection rule for the\nglueball dark matter relic density.\n",
"title": "A Cosmic Selection Rule for Glueball Dark Matter Relic Density"
} | null | null | null | null | true | null | 19878 | null | Default | null | null |
null | {
"abstract": " In this paper, we propose a method using a three dimensional convolutional\nneural network (3-D-CNN) to fuse together multispectral (MS) and hyperspectral\n(HS) images to obtain a high resolution hyperspectral image. Dimensionality\nreduction of the hyperspectral image is performed prior to fusion in order to\nsignificantly reduce the computational time and make the method more robust to\nnoise. Experiments are performed on a data set simulated using a real\nhyperspectral image. The results obtained show that the proposed approach is\nvery promising when compared to conventional methods. This is especially true\nwhen the hyperspectral image is corrupted by additive noise.\n",
"title": "Multispectral and Hyperspectral Image Fusion Using a 3-D-Convolutional Neural Network"
} | null | null | null | null | true | null | 19879 | null | Default | null | null |
null | {
"abstract": " A setting for global variational geometry on Grassmann fibrations is\npresented. The integral variational functionals for finite dimensional immersed\nsubmanifolds are studied by means of the fundamental Lepage equivalent of a\nhomogeneous Lagrangian, which can be regarded as a generalization of the\nwell-known Hilbert form in the classical mechanics. Prolongations of\nimmersions, diffeomorphisms and vector fields to the Grassmann fibrations are\nintroduced as geometric tools for the variations of immersions. The first\ninfinitesimal variation formula together with its consequences, the\nEuler-Lagrange equations for extremal submanifolds and the Noether theorem for\ninvariant variational functionals are proved. The theory is illustrated on the\nvariational functional for minimal submanifolds.\n",
"title": "The fundamental Lepage form in variational theory for submanifolds"
} | null | null | null | null | true | null | 19880 | null | Default | null | null |
null | {
"abstract": " In this paper we revisit some classic board games like Pachisi or the Game of\nGosse. The main contribution of the paper is to design and add some\nfunctionalities to the games in order to transform them in serious games, that\nis, in games with learning and educational purposes. To do that, at the\nbeginning of the game, players choose one or several topics and during the\ngame, players have to anwers questions on these topics in order to move their\nmarkers. We choose classic board games because a lot of people are familiar\nwith them so it is very easy to start to play without wasting time learning\ngame rules and, we think that this is an important element to make the game\nmore attractive to people. To enlarge the number of potential users we have\nimplement the games just using html and javascript and the games can be used in\nany web browser, in any computer (including tablets) , in any computer\narquitecture (Windows, Mac, Linux) and no internet/server conexion is required.\nAssociated software is distributed under Creative Commons\nAttribution-NonCommercial-ShareAlike 3.0 licence and can be obtained at\nthis http URL\n",
"title": "Adding educational funcionalities to classic board games"
} | null | null | null | null | true | null | 19881 | null | Default | null | null |
null | {
"abstract": " In recent years, citizen science has grown in popularity due to a number of\nreasons, including the emphasis on informal learning and creativity potential\nassociated with these initiatives. Citizen science projects address research\nquestions from various domains, ranging from Ecology to Astronomy. Due to the\nadvancement of communication technologies, which makes outreach and engagement\nof wider communities easier, scientists are keen to turn their own research\ninto citizen science projects. However, the development, deployment and\nmanagement of these projects remains challenging. One of the most important\nchallenges is building the project itself. There is no single tool or\nframework, which guides the step-by-step development of the project, since\nevery project has specific characteristics, such as geographical constraints or\nvolunteers' mode of participation. Therefore, in this article, we present a\nseries of conceptual frameworks for categorisation, decision and deployment,\nwhich guide a citizen science project creator in every step of creating a new\nproject starting from the research question to project deployment. The\nframeworks are designed with consideration to the properties of already\nexisting citizen science projects and could be easily extended to include other\ndimensions, which are not currently perceived.\n",
"title": "Conceptual Frameworks for Building Online Citizen Science Projects"
} | null | null | null | null | true | null | 19882 | null | Default | null | null |
null | {
"abstract": " The SRv6 architecture (Segment Routing based on IPv6 data plane) is a\npromising solution to support services like Traffic Engineering, Service\nFunction Chaining and Virtual Private Networks in IPv6 backbones and\ndatacenters. The SRv6 architecture has interesting scalability properties as it\nreduces the amount of state information that needs to be configured in the\nnodes to support the network services. In this paper, we describe the\nadvantages of complementing the SRv6 technology with an SDN based approach in\nbackbone networks. We discuss the architecture of a SRv6 enabled network based\non Linux nodes. In addition, we present the design and implementation of the\nSouthbound API between the SDN controller and the SRv6 device. We have defined\na data-model and four different implementations of the API, respectively based\non gRPC, REST, NETCONF and remote Command Line Interface (CLI). Since it is\nimportant to support both the development and testing aspects we have realized\nan Intent based emulation system to build realistic and reproducible\nexperiments. This collection of tools automate most of the configuration\naspects relieving the experimenter from a significant effort. Finally, we have\nrealized an evaluation of some performance aspects of our architecture and of\nthe different variants of the Southbound APIs and we have analyzed the effects\nof the configuration updates in the SRv6 enabled nodes.\n",
"title": "SDN Architecture and Southbound APIs for IPv6 Segment Routing Enabled Wide Area Networks"
} | null | null | [
"Computer Science"
]
| null | true | null | 19883 | null | Validated | null | null |
null | {
"abstract": " We apply a large-scale computational technique, known as topology\noptimization, to the inverse design of photonic Dirac cones. In particular, we\nreport on a variety of photonic crystal geometries, realizable in simple\nisotropic dielectric materials, which exhibit dual-polarization and\ndual-wavelength Dirac cones. We demonstrate the flexibility of this technique\nby designing photonic crystals of different symmetry types, such as ones with\nfour-fold and six-fold rotational symmetry, which possess Dirac cones at\ndifferent points within the Brillouin zone. The demonstrated and related\noptimization techniques could open new avenues to band-structure engineering\nand manipulating the propagation of light in periodic media, with possible\napplications in exotic optical phenomena such as effective zero-index media and\ntopological photonics.\n",
"title": "Topology-optimized Dual-Polarization Dirac Cones"
} | null | null | null | null | true | null | 19884 | null | Default | null | null |
null | {
"abstract": " Pre-trained word embeddings improve the performance of a neural model at the\ncost of increasing the model size. We propose to benefit from this resource\nwithout paying the cost by operating strictly at the sub-lexical level. Our\napproach is quite simple: before task-specific training, we first optimize\nsub-word parameters to reconstruct pre-trained word embeddings using various\ndistance measures. We report interesting results on a variety of tasks: word\nsimilarity, word analogy, and part-of-speech tagging.\n",
"title": "Reconstruction of Word Embeddings from Sub-Word Parameters"
} | null | null | [
"Computer Science"
]
| null | true | null | 19885 | null | Validated | null | null |
null | {
"abstract": " We start with an overview of a class of submodular functions called SCMMs\n(sums of concave composed with non-negative modular functions plus a final\narbitrary modular). We then define a new class of submodular functions we call\n{\\em deep submodular functions} or DSFs. We show that DSFs are a flexible\nparametric family of submodular functions that share many of the properties and\nadvantages of deep neural networks (DNNs). DSFs can be motivated by considering\na hierarchy of descriptive concepts over ground elements and where one wishes\nto allow submodular interaction throughout this hierarchy. Results in this\npaper show that DSFs constitute a strictly larger class of submodular functions\nthan SCMMs. We show that, for any integer $k>0$, there are $k$-layer DSFs that\ncannot be represented by a $k'$-layer DSF for any $k'<k$. This implies that,\nlike DNNs, there is a utility to depth, but unlike DNNs, the family of DSFs\nstrictly increase with depth. Despite this, we show (using a \"backpropagation\"\nlike method) that DSFs, even with arbitrarily large $k$, do not comprise all\nsubmodular functions. In offering the above results, we also define the notion\nof an antitone superdifferential of a concave function and show how this\nrelates to submodular functions (in general), DSFs (in particular), negative\nsecond-order partial derivatives, continuous submodularity, and concave\nextensions. To further motivate our analysis, we provide various special case\nresults from matroid theory, comparing DSFs with forms of matroid rank, in\nparticular the laminar matroid. Lastly, we discuss strategies to learn DSFs,\nand define the classes of deep supermodular functions, deep difference of\nsubmodular functions, and deep multivariate submodular functions, and discuss\nwhere these can be useful in applications.\n",
"title": "Deep Submodular Functions"
} | null | null | null | null | true | null | 19886 | null | Default | null | null |
null | {
"abstract": " We consider the inverse problems of determining the potential or the damping\ncoefficient appearing in the wave equation. We will prove the unique\ndetermination of these coefficients from the one point measurement. Since our\nproblem is under-determined, so some extra assumption on the coefficients is\nrequired to prove the uniqueness.\n",
"title": "Inverse problems for the wave equation with under-determined data"
} | null | null | null | null | true | null | 19887 | null | Default | null | null |
null | {
"abstract": " A tragedy of the commons (TOC) occurs when individuals acting in their own\nself-interest deplete commonly-held resources, leading to a worse outcome than\nhad they cooperated. Over time, the depletion of resources can change\nincentives for subsequent actions. Here, we investigate long-term feedback\nbetween game and environment across a continuum of incentives in an\nindividual-based framework. We identify payoff-dependent transition rules that\nlead to oscillatory TOC-s in stochastic simulations and the mean field limit.\nFurther extending the stochastic model, we find that spatially explicit\ninteractions can lead to emergent, localized dynamics, including the\npropagation of cooperative wave fronts and cluster formation of both social\ncontext and resources. These dynamics suggest new mechanisms underlying how\nTOCs arise and how they might be averted.\n",
"title": "Spatial interactions and oscillatory tragedies of the commons"
} | null | null | null | null | true | null | 19888 | null | Default | null | null |
null | {
"abstract": " The linear growth of operators in local quantum systems leads to an effective\nlightcone even if the system is non-relativistic. We show that consistency of\ndiffusive transport with this lightcone places an upper bound on the\ndiffusivity: $D \\lesssim v^2 \\tau_\\text{eq}$. The operator growth velocity $v$\ndefines the lightcone and $\\tau_\\text{eq}$ is the local equilibration\ntimescale, beyond which the dynamics of conserved densities is diffusive. We\nverify that the bound is obeyed in various weakly and strongly interacting\ntheories. In holographic models this bound establishes a relation between the\nhydrodynamic and leading non-hydrodynamic quasinormal modes of planar black\nholes. Our bound relates transport data --- including the electrical\nresistivity and the shear viscosity --- to the local equilibration time, even\nin the absence of a quasiparticle description. In this way, the bound sheds\nlight on the observed $T$-linear resistivity of many unconventional metals, the\nshear viscosity of the quark-gluon plasma and the spin transport of unitary\nfermions.\n",
"title": "An upper bound on transport"
} | null | null | null | null | true | null | 19889 | null | Default | null | null |
null | {
"abstract": " Each Multiplicative Exponential Linear Logic (MELL) proof-net can be expanded\ninto a differential net, which is its Taylor expansion. We prove that two\ndifferent MELL proof-nets have two different Taylor expansions. As a corollary,\nwe prove a completeness result for MELL: We show that the relational model is\ninjective for MELL proof-nets, i.e. the equality between MELL proof-nets in the\nrelational model is exactly axiomatized by cut-elimination.\n",
"title": "Taylor expansion in linear logic is invertible"
} | null | null | null | null | true | null | 19890 | null | Default | null | null |
null | {
"abstract": " Multi robot systems have the potential to be utilized in a variety of\napplications. In most of the previous works, the trajectory generation for\nmulti robot systems is implemented in known environments. To overcome that we\npresent an online trajectory optimization algorithm that utilizes communication\nof robots' current states to account to the other robots while using local\nobject based maps for identifying obstacles. Based upon this data, we predict\nthe trajectory expected to be traversed by the robots and utilize that to avoid\ncollisions by formulating regions of free space that the robot can be without\ncolliding with other robots and obstacles. A trajectory is optimized\nconstraining the robot to remain within this region.The proposed method is\ntested in simulations on Gazebo using ROS.\n",
"title": "Collision-Free Multi Robot Trajectory Optimization in Unknown Environments using Decentralized Trajectory Planning"
} | null | null | null | null | true | null | 19891 | null | Default | null | null |
null | {
"abstract": " For a Riemannian covering $M_1\\to M_0$ of complete Riemannian manifolds with\nboundary (possibly empty) and respective fundamental groups\n$\\Gamma_1\\subseteq\\Gamma_0$, we show that the bottoms of the spectra of $M_0$\nand $M_1$ coincide if the right action of $\\Gamma_0$ on\n$\\Gamma_1\\backslash\\Gamma_0$ is amenable.\n",
"title": "On the bottom of spectra under coverings"
} | null | null | null | null | true | null | 19892 | null | Default | null | null |
null | {
"abstract": " We extend the theory of ground states of classical Heisenberg spin systems\npreviously published to the case where the interaction with an external\nmagnetic field is described by a Zeeman term. The ground state problem for the\nHeisenberg-Zeeman Hamiltonian can be reduced first to the relative ground state\nproblem, and, in a second step, to the absolute ground state problem for pure\nHeisenberg Hamiltonians depending on an additional Lagrange parameter. We\ndistinguish between continuous and discontinuous reduction. Moreover, there are\nvarious general statements about Heisenberg-Zeeman systems that will be proven\nunder most general assumptions. One topic is the connection between the minimal\nenergy functions $E_{min}$ for the Heisenberg energy and $H_{min}$ for the\nHeisenberg-Zeeman energy which turn out to be essentially mutual\nLegendre-Fenchel transforms. This generalization of the traditional Legendre\ntransform is especially suited to cope with situations where the function\n$E_{min}$ is not convex and consequently there is a magnetization jump at a\ncritical field. Another topic is magnetization and the occurrence of threshold\nfields $B_{thr}$ and saturation fields $B_{sat}$, where we provide a general\nformula for the latter. We suggest a distinction between ferromagnetic and\nanti-ferromagnetic systems based on the vanishing of $B_{sat}$ for the former\nones. Parabolic systems are defined in such a way that $E_{min}$ and $H_{min}$\nhave a particularly simple form and studied in detail. For a large class of\nparabolic systems the relative ground states can be constructed from the\nabsolute ground state by means of a so-called umbrella family. Finally we\nprovide a counter-example of a parabolic system where this construction is not\npossible.\n",
"title": "Theory of ground states for classical Heisenberg spin systems IV"
} | null | null | null | null | true | null | 19893 | null | Default | null | null |
null | {
"abstract": " We present a framework for vision-based model predictive control (MPC) for\nthe task of aggressive, high-speed autonomous driving. Our approach uses deep\nconvolutional neural networks to predict cost functions from input video which\nare directly suitable for online trajectory optimization with MPC. We\ndemonstrate the method in a high speed autonomous driving scenario, where we\nuse a single monocular camera and a deep convolutional neural network to\npredict a cost map of the track in front of the vehicle. Results are\ndemonstrated on a 1:5 scale autonomous vehicle given the task of high speed,\naggressive driving.\n",
"title": "Aggressive Deep Driving: Model Predictive Control with a CNN Cost Model"
} | null | null | null | null | true | null | 19894 | null | Default | null | null |
null | {
"abstract": " We give a $K$-theoretic criterion for a quasi-projective variety to be\nsmooth. If $\\mathbb{L}$ is a line bundle corresponding to an ample invertible\nsheaf on $X$, it suffices that $K_q(X) = K_q(\\mathbb{L})$ for all\n$q\\le\\dim(X)+1$.\n",
"title": "K-theory of line bundles and smooth varieties"
} | null | null | null | null | true | null | 19895 | null | Default | null | null |
null | {
"abstract": " Modern surveys have provided the astronomical community with a flood of\nhigh-dimensional data, but analyses of these data often occur after their\nprojection to lower-dimensional spaces. In this work, we introduce a local\ntwo-sample hypothesis test framework that an analyst may directly apply to data\nin their native space. In this framework, the analyst defines two classes based\non a response variable of interest (e.g. higher-mass galaxies versus lower-mass\ngalaxies) and determines at arbitrary points in predictor space whether the\nlocal proportions of objects that belong to the two classes significantly\ndiffers from the global proportion.\nOur framework has a potential myriad of uses throughout astronomy; here, we\ndemonstrate its efficacy by applying it to a sample of 2487 i-band-selected\ngalaxies observed by the HST ACS in four of the CANDELS program fields. For\neach galaxy, we have seven morphological summary statistics along with an\nestimated stellar mass and star-formation rate. We perform two studies: one in\nwhich we determine regions of the seven-dimensional space of morphological\nstatistics where high-mass galaxies are significantly more numerous than\nlow-mass galaxies, and vice-versa, and another study where we use SFR in place\nof mass. We find that we are able to identify such regions, and show how\nhigh-mass/low-SFR regions are associated with concentrated and undisturbed\ngalaxies while galaxies in low-mass/high-SFR regions appear more extended\nand/or disturbed than their high-mass/low-SFR counterparts.\n",
"title": "Local Two-Sample Testing: A New Tool for Analysing High-Dimensional Astronomical Data"
} | null | null | null | null | true | null | 19896 | null | Default | null | null |
null | {
"abstract": " We propose a novel approach towards adversarial attacks on neural networks\n(NN), focusing on tampering the data used for training instead of generating\nattacks on trained models. Our network-agnostic method creates a backdoor\nduring training which can be exploited at test time to force a neural network\nto exhibit abnormal behaviour. We demonstrate on two widely used datasets\n(CIFAR-10 and SVHN) that a universal modification of just one pixel per image\nfor all the images of a class in the training set is enough to corrupt the\ntraining procedure of several state-of-the-art deep neural networks causing the\nnetworks to misclassify any images to which the modification is applied. Our\naim is to bring to the attention of the machine learning community, the\npossibility that even learning-based methods that are personally trained on\npublic datasets can be subject to attacks by a skillful adversary.\n",
"title": "Are You Tampering With My Data?"
} | null | null | null | null | true | null | 19897 | null | Default | null | null |
null | {
"abstract": " We show how to answer spatial multiple-set intersection queries in O(n(log\nw)/w + kt) expected time, where n is the total size of the t sets involved in\nthe query, w is the number of bits in a memory word, k is the output size, and\nc is any fixed constant. This improves the asymptotic performance over previous\nsolutions and is based on an interesting data structure, known as 2-3 cuckoo\nhash-filters. Our results apply in the word-RAM model (or practical RAM model),\nwhich allows for constant-time bit-parallel operations, such as bitwise AND,\nOR, NOT, and MSB (most-significant 1-bit), as exist in modern CPUs and GPUs.\nOur solutions apply to any multiple-set intersection queries in spatial data\nsets that can be reduced to one-dimensional range queries, such as spatial join\nqueries for one-dimensional points or sets of points stored along space-filling\ncurves, which are used in GIS applications.\n",
"title": "Answering Spatial Multiple-Set Intersection Queries Using 2-3 Cuckoo Hash-Filters"
} | null | null | null | null | true | null | 19898 | null | Default | null | null |
null | {
"abstract": " We present a framework to calculate the cascade size evolution for a large\nclass of cascade models on random network ensembles in the limit of infinite\nnetwork size. Our method is exact and applies to network ensembles with almost\narbitrary degree distribution, degree-degree correlations and, in case of\nthreshold models, with arbitrary threshold distribution. With our approach, we\nshift the perspective from the known branching process approximations to the\niterative update of suitable probability distributions. Such distributions are\nkey to capture cascade dynamics that involve possibly continuous quantities and\nthat depend on the cascade history, e.g. if load is accumulated over time.\nThese distributions respect the Markovian nature of the studied random\nprocesses. Random variables capture the impact of nodes that have failed at any\npoint in the past on their neighborhood. As a proof of concept, we provide two\nexamples: (a) Constant load models that cover many of the analytically\ntractable cascade models, and, as a highlight, (b) a fiber bundle model that\nwas not tractable by branching process approximations before. Our derivations\ncover the whole cascade dynamics, not only their steady state. This allows to\ninclude interventions in time or further model complexity in the analysis.\n",
"title": "A framework for cascade size calculations on random networks"
} | null | null | null | null | true | null | 19899 | null | Default | null | null |
null | {
"abstract": " Monge-Kantorovich distances, otherwise known as Wasserstein distances, have\nreceived a growing attention in statistics and machine learning as a powerful\ndiscrepancy measure for probability distributions. In this paper, we focus on\nforecasting a Gaussian process indexed by probability distributions. For this,\nwe provide a family of positive definite kernels built using transportation\nbased distances. We provide a probabilistic understanding of these kernels and\ncharacterize the corresponding stochastic processes. We prove that the Gaussian\nprocesses indexed by distributions corresponding to these kernels can be\nefficiently forecast, opening new perspectives in Gaussian process modeling.\n",
"title": "A Gaussian Process Regression Model for Distribution Inputs"
} | null | null | null | null | true | null | 19900 | null | Default | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.