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": " Quantum technology is increasingly relying on specialised statistical\ninference methods for analysing quantum measurement data. This motivates the\ndevelopment of \"quantum statistics\", a field that is shaping up at the overlap\nof quantum physics and \"classical\" statistics. One of the less investigated\ntopics to date is that of statistical inference for infinite dimensional\nquantum systems, which can be seen as quantum counterpart of non-parametric\nstatistics. In this paper we analyse the asymptotic theory of quantum\nstatistical models consisting of ensembles of quantum systems which are\nidentically prepared in a pure state. In the limit of large ensembles we\nestablish the local asymptotic equivalence (LAE) of this i.i.d. model to a\nquantum Gaussian white noise model. We use the LAE result in order to establish\nminimax rates for the estimation of pure states belonging to Hermite-Sobolev\nclasses of wave functions. Moreover, for quadratic functional estimation of the\nsame states we note an elbow effect in the rates, whereas for testing a pure\nstate a sharp parametric rate is attained over the nonparametric\nHermite-Sobolev class.\n",
"title": "Local asymptotic equivalence of pure quantum states ensembles and quantum Gaussian white noise"
}
| null | null | null | null | true | null |
14101
| null |
Default
| null | null |
null |
{
"abstract": " The wide adoption of DNNs has given birth to unrelenting computing\nrequirements, forcing datacenter operators to adopt domain-specific\naccelerators to train them. These accelerators typically employ densely packed\nfull precision floating-point arithmetic to maximize performance per area.\nOngoing research efforts seek to further increase that performance density by\nreplacing floating-point with fixed-point arithmetic. However, a significant\nroadblock for these attempts has been fixed point's narrow dynamic range, which\nis insufficient for DNN training convergence. We identify block floating point\n(BFP) as a promising alternative representation since it exhibits wide dynamic\nrange and enables the majority of DNN operations to be performed with\nfixed-point logic. Unfortunately, BFP alone introduces several limitations that\npreclude its direct applicability. In this work, we introduce HBFP, a hybrid\nBFP-FP approach, which performs all dot products in BFP and other operations in\nfloating point. HBFP delivers the best of both worlds: the high accuracy of\nfloating point at the superior hardware density of fixed point. For a wide\nvariety of models, we show that HBFP matches floating point's accuracy while\nenabling hardware implementations that deliver up to 8.5x higher throughput.\n",
"title": "Training DNNs with Hybrid Block Floating Point"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14102
| null |
Validated
| null | null |
null |
{
"abstract": " Spectral clustering has found extensive use in many areas. Most traditional\nspectral clustering algorithms work in three separate steps: similarity graph\nconstruction; continuous labels learning; discretizing the learned labels by\nk-means clustering. Such common practice has two potential flaws, which may\nlead to severe information loss and performance degradation. First, predefined\nsimilarity graph might not be optimal for subsequent clustering. It is\nwell-accepted that similarity graph highly affects the clustering results. To\nthis end, we propose to automatically learn similarity information from data\nand simultaneously consider the constraint that the similarity matrix has exact\nc connected components if there are c clusters. Second, the discrete solution\nmay deviate from the spectral solution since k-means method is well-known as\nsensitive to the initialization of cluster centers. In this work, we transform\nthe candidate solution into a new one that better approximates the discrete\none. Finally, those three subtasks are integrated into a unified framework,\nwith each subtask iteratively boosted by using the results of the others\ntowards an overall optimal solution. It is known that the performance of a\nkernel method is largely determined by the choice of kernels. To tackle this\npractical problem of how to select the most suitable kernel for a particular\ndata set, we further extend our model to incorporate multiple kernel learning\nability. Extensive experiments demonstrate the superiority of our proposed\nmethod as compared to existing clustering approaches.\n",
"title": "Unified Spectral Clustering with Optimal Graph"
}
| null | null | null | null | true | null |
14103
| null |
Default
| null | null |
null |
{
"abstract": " Measurements of high-velocity clouds' metallicities provide important clues\nabout their origins, and hence on whether they play a role in fueling ongoing\nstar formation in the Galaxy. However, accurate interpretation of these\nmeasurements requires compensating for the galactic material that has been\nmixed into the clouds. In order to determine how much the metallicity changes\nas a result of this mixing, we have carried out three-dimensional\nwind-tunnel-like hydrodynamical simulations of an example cloud. Our model\ncloud is patterned after the Smith Cloud, a particularly well-studied cloud of\nmass $\\sim 5 \\times 10^6~M_\\odot$. We calculated the fraction of the\nhigh-velocity material that had originated in the galactic halo,\n$F_\\mathrm{h}$, for various sight lines passing through our model cloud. We\nfind that $F_\\mathrm{h}$ generally increases with distance from the head of the\ncloud, reaching $\\sim$0.5 in the tail of the cloud. Models in which the\nmetallicities (relative to solar) of the original cloud, $Z_\\mathrm{cl}$, and\nof the halo, $Z_\\mathrm{h}$, are in the approximate ranges $0.1 \\lesssim\nZ_\\mathrm{cl} \\lesssim 0.3$ and $0.7 \\lesssim Z_\\mathrm{h} \\lesssim 1.0$,\nrespectively, are in rough agreement with the observations. Models with\n$Z_\\mathrm{h} \\sim 0.1$ and $Z_\\mathrm{cl} \\gtrsim 0.5$ are also in rough\nagreement with the observations, but such a low halo metallicity is\ninconsistent with recent independent measurements. We conclude that the Smith\nCloud's observed metallicity may not be a true reflection of its original\nmetallicity and that the cloud's ultimate origin remains uncertain.\n",
"title": "The Effect of Mixing on the Observed Metallicity of the Smith Cloud"
}
| null | null | null | null | true | null |
14104
| null |
Default
| null | null |
null |
{
"abstract": " Deep generative models trained with large amounts of unlabelled data have\nproven to be powerful within the domain of unsupervised learning. Many real\nlife data sets contain a small amount of labelled data points, that are\ntypically disregarded when training generative models. We propose the\nCluster-aware Generative Model, that uses unlabelled information to infer a\nlatent representation that models the natural clustering of the data, and\nadditional labelled data points to refine this clustering. The generative\nperformances of the model significantly improve when labelled information is\nexploited, obtaining a log-likelihood of -79.38 nats on permutation invariant\nMNIST, while also achieving competitive semi-supervised classification\naccuracies. The model can also be trained fully unsupervised, and still improve\nthe log-likelihood performance with respect to related methods.\n",
"title": "Semi-Supervised Generation with Cluster-aware Generative Models"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14105
| null |
Validated
| null | null |
null |
{
"abstract": " We present a generalization bound for feedforward neural networks in terms of\nthe product of the spectral norm of the layers and the Frobenius norm of the\nweights. The generalization bound is derived using a PAC-Bayes analysis.\n",
"title": "A PAC-Bayesian Approach to Spectrally-Normalized Margin Bounds for Neural Networks"
}
| null | null | null | null | true | null |
14106
| null |
Default
| null | null |
null |
{
"abstract": " In the preceding paper (Efroimsky 2017), we derived an expression for the\ntidal dissipation rate in a homogeneous near-spherical Maxwell body librating\nin longitude. Now, by equating this expression to the outgoing energy flux due\nto the vapour plumes, we estimate the mean tidal viscosity of Enceladus, under\nthe assumption that the Enceladean mantle behaviour is Maxwell. This method\nyields a value of $\\,0.24\\times 10^{14}\\;\\mbox{Pa~s}\\,$ for the mean tidal\nviscosity, which is very close to the viscosity of ice near the melting point.\n",
"title": "Tidal viscosity of Enceladus"
}
| null | null |
[
"Physics"
] | null | true | null |
14107
| null |
Validated
| null | null |
null |
{
"abstract": " In this work, we introduce the MOldavian and ROmanian Dialectal COrpus\n(MOROCO), which is freely available for download at\nthis https URL. The corpus contains 33564 samples of\ntext (with over 10 million tokens) collected from the news domain. The samples\nbelong to one of the following six topics: culture, finance, politics, science,\nsports and tech. The data set is divided into 21719 samples for training, 5921\nsamples for validation and another 5924 samples for testing. For each sample,\nwe provide corresponding dialectal and category labels. This allows us to\nperform empirical studies on several classification tasks such as (i) binary\ndiscrimination of Moldavian versus Romanian text samples, (ii) intra-dialect\nmulti-class categorization by topic and (iii) cross-dialect multi-class\ncategorization by topic. We perform experiments using a shallow approach based\non string kernels, as well as a novel deep approach based on character-level\nconvolutional neural networks containing Squeeze-and-Excitation blocks. We also\npresent and analyze the most discriminative features of our best performing\nmodel, before and after named entity removal.\n",
"title": "MOROCO: The Moldavian and Romanian Dialectal Corpus"
}
| null | null | null | null | true | null |
14108
| null |
Default
| null | null |
null |
{
"abstract": " From the Einstein field equations, in a weak-field approximation and for\nspeeds small compared to the speed of light in vacuum, the following system is\nobtained \\begin{align*}\n\\nabla \\times \\overrightarrow{E_g} & =\n-\\frac{1}{c} \\frac{\\partial \\overrightarrow{B_g}}{\\partial t},\n\\nabla \\cdot \\overrightarrow{E_g} \\;\\; & \\approx -4\\pi G\\rho_g,\n\\nabla \\times \\overrightarrow{B_g} & \\approx\n-\\frac{4\\pi G}{c^{2}}\\overrightarrow{J_g}+\n\\frac{1}{c}\\frac{\\partial \\overrightarrow{E_g}}{\\partial t},\n\\nabla \\cdot \\overrightarrow{B_g} \\;\\; & = 0, \\end{align*} where\n$\\overrightarrow{E_g}$ is the gravitoelectric field, $\\overrightarrow{B_g}$ is\nthe gravitomagnetic field, $\\overrightarrow{J_g}$ is the space-time-mass\ncurrent density and $\\rho_g$ is the space-time-mass density. This last\ngravitoelectromagnetic field system is similar to the Maxwell equations, thus\nshowing an analogy between the electromagnetic theory and gravitation.\n",
"title": "Linearized Einstein's field equations"
}
| null | null | null | null | true | null |
14109
| null |
Default
| null | null |
null |
{
"abstract": " We propose a novel tree classification system called Treelogy, that fuses\ndeep representations with hand-crafted features obtained from leaf images to\nperform leaf-based plant classification. Key to this system are segmentation of\nthe leaf from an untextured background, using convolutional neural networks\n(CNNs) for learning deep representations, extracting hand-crafted features with\na number of image processing techniques, training a linear SVM with feature\nvectors, merging SVM and CNN results, and identifying the species from a\ndataset of 57 trees. Our classification results show that fusion of deep\nrepresentations with hand-crafted features leads to the highest accuracy. The\nproposed algorithm is embedded in a smart-phone application, which is publicly\navailable. Furthermore, our novel dataset comprised of 5408 leaf images is also\nmade public for use of other researchers.\n",
"title": "Treelogy: A Novel Tree Classifier Utilizing Deep and Hand-crafted Representations"
}
| null | null | null | null | true | null |
14110
| null |
Default
| null | null |
null |
{
"abstract": " We investigate flow instability created by an oblique shock wave impinging on\na Mach 5.92 laminar boundary layer at a transitional Reynolds number. The\nadverse pressure gradient of the oblique shock causes the boundary layer to\nseparate from the wall, resulting in the formation of a recirculation bubble.\nFor sufficiently large oblique shock angles, the recirculation bubble is\nunstable to three-dimensional perturbations and the flow bifurcates from its\noriginal laminar state. We utilize Direct Numerical Simulation (DNS) and Global\nStability Analysis (GSA) to show that this first occurs at a critical shock\nangle of $\\theta = 12.9^o$. At bifurcation, the least stable global mode is\nnon-oscillatory, and it takes place at a spanwise wavenumber $\\beta=0.25$, in\ngood agreement with DNS results. Examination of the critical global mode\nreveals that it originates from an interaction between small spanwise\ncorrugations at the base of the incident shock, streamwise vortices inside the\nrecirculation bubble, and spanwise modulation of the bubble strength. The\nglobal mode drives the formation of long streamwise streaks downstream of the\nbubble. While the streaks may be amplified by either the lift-up effect or by\nGörtler instability, we show that centrifugal instability plays no role in\nthe upstream self-sustaining mechanism of the global mode. We employ an adjoint\nsolver to corroborate our physical interpretation by showing that the critical\nglobal mode is most sensitive to base flow modifications that are entirely\ncontained inside the recirculation bubble.\n",
"title": "Simulation and stability analysis of oblique shock wave/boundary layer interactions at Mach 5.92"
}
| null | null |
[
"Physics"
] | null | true | null |
14111
| null |
Validated
| null | null |
null |
{
"abstract": " The paper gives an introduction to rate equations in nonlinear continuum\nmechanics which should obey specific transformation rules. Emphasis is placed\non the geometrical nature of the operations involved in order to clarify the\ndifferent concepts. The paper is particularly concerned with common classes of\nconstitutive equations based on corotational stress rates and their proper\nimplementation in time for solving initial boundary value problems. Hypoelastic\nsimple shear is considered as an example application for the derived theory and\nalgorithms.\n",
"title": "Notes on rate equations in nonlinear continuum mechanics"
}
| null | null | null | null | true | null |
14112
| null |
Default
| null | null |
null |
{
"abstract": " For many modern applications in science and engineering, data are collected\nin a streaming fashion carrying time-varying information, and practitioners\nneed to process them with a limited amount of memory and computational\nresources in a timely manner for decision making. This often is coupled with\nthe missing data problem, such that only a small fraction of data attributes\nare observed. These complications impose significant, and unconventional,\nconstraints on the problem of streaming Principal Component Analysis (PCA) and\nsubspace tracking, which is an essential building block for many inference\ntasks in signal processing and machine learning. This survey article reviews a\nvariety of classical and recent algorithms for solving this problem with low\ncomputational and memory complexities, particularly those applicable in the big\ndata regime with missing data. We illustrate that streaming PCA and subspace\ntracking algorithms can be understood through algebraic and geometric\nperspectives, and they need to be adjusted carefully to handle missing data.\nBoth asymptotic and non-asymptotic convergence guarantees are reviewed.\nFinally, we benchmark the performance of several competitive algorithms in the\npresence of missing data for both well-conditioned and ill-conditioned systems.\n",
"title": "Streaming PCA and Subspace Tracking: The Missing Data Case"
}
| null | null | null | null | true | null |
14113
| null |
Default
| null | null |
null |
{
"abstract": " In this study, we address the question whether (and to what extent,\nrespectively) altmetrics are related to the scientific quality of papers (as\nmeasured by peer assessments). Only a few studies have previously investigated\nthe relationship between altmetrics and assessments by peers. In the first\nstep, we analyse the underlying dimensions of measurement for traditional\nmetrics (citation counts) and altmetrics - by using principal component\nanalysis (PCA) and factor analysis (FA). In the second step, we test the\nrelationship between the dimensions and quality of papers (as measured by the\npost-publication peer-review system of F1000Prime assessments) - using\nregression analysis. The results of the PCA and FA show that altmetrics operate\nalong different dimensions, whereas Mendeley counts are related to citation\ncounts, and tweets form a separate dimension. The results of the regression\nanalysis indicate that citation-based metrics and readership counts are\nsignificantly more related to quality, than tweets. This result on the one hand\nquestions the use of Twitter counts for research evaluation purposes and on the\nother hand indicates potential use of Mendeley reader counts.\n",
"title": "Do altmetrics correlate with the quality of papers? A large-scale empirical study based on F1000Prime data"
}
| null | null | null | null | true | null |
14114
| null |
Default
| null | null |
null |
{
"abstract": " The declination is a quantitative method for identifying possible partisan\ngerrymanders by analyzing vote distributions. In this expository note we\nexplain and motivate the definition of the declination. The minimal computer\ncode required for computing the declination is included. We end by computing\nits value on several recent elections.\n",
"title": "Introduction to the declination function for gerrymanders"
}
| null | null | null | null | true | null |
14115
| null |
Default
| null | null |
null |
{
"abstract": " Dynamic adaptive streaming over HTTP (DASH) has recently been widely deployed\nin the Internet and adopted in the industry. It, however, does not impose any\nadaptation logic for selecting the quality of video fragments requested by\nclients and suffers from lackluster performance with respect to a number of\ndesirable properties: efficiency, stability, and fairness when multiple players\ncompete for a bottleneck link. In this paper, we propose a throughput-friendly\nDASH (TFDASH) rate control scheme for video streaming with multiple clients\nover DASH to well balance the trade-offs among efficiency, stability, and\nfairness. The core idea behind guaranteeing fairness and high efficiency\n(bandwidth utilization) is to avoid OFF periods during the downloading process\nfor all clients, i.e., the bandwidth is in perfect-subscription or\nover-subscription with bandwidth utilization approach to 100\\%. We also propose\na dual-threshold buffer model to solve the instability problem caused by the\nabove idea. As a result, by integrating these novel components, we also propose\na probability-driven rate adaption logic taking into account several key\nfactors that most influence visual quality, including buffer occupancy, video\nplayback quality, video bit-rate switching frequency and amplitude, to\nguarantee high-quality video streaming. Our experiments evidently demonstrate\nthe superior performance of the proposed method.\n",
"title": "TFDASH: A Fairness, Stability, and Efficiency Aware Rate Control Approach for Multiple Clients over DASH"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14116
| null |
Validated
| null | null |
null |
{
"abstract": " Evolution and propagation of the world's languages is a complex phenomenon,\ndriven, to a large extent, by social interactions. Multilingual society can be\nseen as a system of interacting agents, where the interaction leads to a\nmodification of the language spoken by the individuals. Two people can reach\nthe state of full linguistic compatibility due to the positive interactions,\nlike transfer of loanwords. But, on the other hand, if they speak entirely\ndifferent languages, they will separate from each other. These simple\nobservations make the network science the most suitable framework to describe\nand analyze dynamics of language change. Although many mechanisms have been\nexplained, we lack a qualitative description of the scaling behavior for\ndifferent sizes of a population. Here we address the issue of the language\ndiversity in societies of different sizes, and we show that local interactions\nare crucial to capture characteristics of the empirical data. We propose a\nmodel of social interactions, extending the idea from, that explains the growth\nof the language diversity with the size of a population of country or society.\nWe argue that high clustering and network disintegration are the most important\ncharacteristics of models properly describing empirical data. Furthermore, we\ncancel the contradiction between previous models and the Solomon Islands case.\nOur results demonstrate the importance of the topology of the network, and the\nrewiring mechanism in the process of language change.\n",
"title": "Predicting language diversity with complex network"
}
| null | null | null | null | true | null |
14117
| null |
Default
| null | null |
null |
{
"abstract": " This article is dedicated to the late Giorgio Israel. R{é}sum{é}. The aim\nof this article is to propose on the one hand a brief history of modeling\nstarting from the works of Fibonacci, Robert Malthus, Pierre Francis Verhulst\nand then Vito Volterra and, on the other hand, to present the main hypotheses\nof the very famous but very little known predator-prey model elaborated in the\n1920s by Volterra in order to solve a problem posed by his son-in-law, Umberto\nD'Ancona. It is thus shown that, contrary to a widely-held notion, Volterra's\nmodel is realistic and his seminal work laid the groundwork for modern\npopulation dynamics and mathematical ecology, including seasonality, migration,\npollution and more. 1. A short history of modeling 1.1. The Malthusian model.\nIf the rst scientic view of population growth seems to be that of Leonardo\nFibonacci [2], also called Leonardo of Pisa, whose famous sequence of numbers\nwas presented in his Liber abaci (1202) as a solution to a population growth\nproblem, the modern foundations of population dynamics clearly date from Thomas\nRobert Malthus [20]. Considering an ideal population consisting of a single\nhomogeneous animal species, that is, neglecting the variations in age, size and\nany periodicity for birth or mortality, and which lives alone in an invariable\nenvironment or coexists with other species without any direct or indirect\ninuence, he founded in 1798, with his celebrated claim Population, when\nunchecked, increases in a geometrical ratio, the paradigm of exponential\ngrowth. This consists in assuming that the increase of the number N (t) of\nindividuals of this population, during a short interval of time, is\nproportional to N (t). This translates to the following dierential equation :\n(1) dN (t) dt = $\\epsilon$N (t) where $\\epsilon$ is a constant factor of\nproportionality that represents the growth coe-cient or growth rate. By\nintegrating (1) we obtain the law of exponential growth or law of Malthusian\ngrowth (see Fig. 1). This law, which does not take into account the limits\nimposed by the environment on growth and which is in disagreement with the\nactual facts, had a profound inuence on Charles Darwin's work on natural\nselection. Indeed, Darwin [1] founded the idea of survival of the ttest on the\n1. According to Frontier and Pichod-Viale [3] the correct terminology should be\npopulation kinetics, since the interaction between species cannot be\nrepresented by forces. 2. A population is dened as the set of individuals of\nthe same species living on the same territory and able to reproduce among\nthemselves.\n",
"title": "The paradox of Vito Volterra's predator-prey model"
}
| null | null | null | null | true | null |
14118
| null |
Default
| null | null |
null |
{
"abstract": " A categorical point of view about minimization in subrecursive classes is\npresented by extending the concept of Symmetric Monoidal Comprehension to that\nof Distributive Minimization Comprehension. This is achieved by endowing the\nformer with coproducts and a finality condition for coalgebras over the\nendofunctor sending X to ${1}\\oplus{X}$ to perform a safe minimization\noperator. By relying on the characterization given by Bellantoni, a tiered\nstructure is presented from which one can obtain the levels of the Polytime\nHierarchy as those classes of partial functions obtained after a certain number\nof minimizations.\n",
"title": "Distributive Minimization Comprehensions and the Polynomial Hierarchy"
}
| null | null | null | null | true | null |
14119
| null |
Default
| null | null |
null |
{
"abstract": " Devaney and Krych showed that for the exponential family $\\lambda e^z$, where\n$0<\\lambda <1/e$, the Julia set consists of uncountably many pairwise disjoint\nsimple curves tending to $\\infty$. Viana proved that these curves are smooth.\nIn this article we consider a quasiregular counterpart of the exponential map,\nthe so-called Zorich maps, and generalize Viana's result to these maps.\n",
"title": "On the differentiability of hairs for Zorich maps"
}
| null | null | null | null | true | null |
14120
| null |
Default
| null | null |
null |
{
"abstract": " Motivated by ride-sharing platforms' efforts to reduce their riders' wait\ntimes for a vehicle, this paper introduces a novel problem of placing vehicles\nto fulfill real-time pickup requests in a spatially and temporally changing\nenvironment. The real-time nature of this problem makes it fundamentally\ndifferent from other placement and scheduling problems, as it requires not only\nreal-time placement decisions but also handling real-time request dynamics,\nwhich are influenced by human mobility patterns. We use a dataset of ten\nmillion ride requests from four major U.S. cities to show that the requests\nexhibit significant self-similarity. We then propose distributed online\nlearning algorithms for the real-time vehicle placement problem and bound their\nexpected performance under this observed self-similarity.\n",
"title": "On the Real-time Vehicle Placement Problem"
}
| null | null | null | null | true | null |
14121
| null |
Default
| null | null |
null |
{
"abstract": " Modern systems will increasingly rely on energy harvested from their\nenvironment. Such systems utilize batteries to smoothen out the random\nfluctuations in harvested energy. These fluctuations induce highly variable\nbattery charge and discharge rates, which affect the efficiencies of practical\nbatteries that typically have non-zero internal resistances. In this paper, we\nstudy an energy harvesting communication system using a finite battery with\nnon-zero internal resistance. We adopt a dual-path architecture, in which\nharvested energy can be directly used, or stored and then used. In a frame,\nboth time and power can be split between energy storage and data transmission.\nFor a single frame, we derive an analytical expression for the rate optimal\ntime and power splitting ratios between harvesting energy and transmitting\ndata. We then optimize the time and power splitting ratios for a group of\nframes, assuming non-causal knowledge of harvested power and fading channel\ngains, by giving an approximate solution. When only the statistics of the\nenergy arrivals and channel gains are known, we derive a dynamic programming\nbased policy and, propose three sub-optimal policies, which are shown to\nperform competitively. In summary, our study suggests that battery internal\nresistance significantly impacts the design and performance of energy\nharvesting communication systems and must be taken into account.\n",
"title": "Energy Harvesting Communication Using Finite-Capacity Batteries with Internal Resistance"
}
| null | null | null | null | true | null |
14122
| null |
Default
| null | null |
null |
{
"abstract": " The evolution of smart microgrid and its demand-response characteristics not\nonly will change the paradigms of the century-old electric grid but also will\nshape the electricity market. In this new market scenario, once always energy\nconsumers, now may act as sellers due to the excess of energy generated from\nnewly deployed distributed generators (DG). The smart microgrid will use the\nexisting electrical transmission network and a pay per use transportation cost\nwithout implementing new transmission lines which involve a massive capital\ninvestment. In this paper, we propose a novel algorithm to minimize the\nelectricity price with the optimal trading of energy between sellers and buyers\nof the smart microgrid network. The algorithm is capable of solving the optimal\npower allocation problem (with optimal transmission cost) for a microgrid\nnetwork in a polynomial time without modifying the actual marginal costs of\npower generation. We mathematically formulate the problem as a nonlinear\nnon-convex and decompose the problem to separate the optimal marginal cost\nmodel from the electricity allocation model. Then, we develop a\ndivide-and-conquer method to minimize the electricity price by jointly solving\nthe optimal marginal cost model and electricity allocation problems. To\nevaluate the performance of the solution method, we develop and simulate the\nmodel with different marginal cost functions and compare it with a first come\nfirst serve electricity allocation method.\n",
"title": "A Novel Algorithm for Optimal Electricity Pricing in a Smart Microgrid Network"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14123
| null |
Validated
| null | null |
null |
{
"abstract": " Our premise is that autonomous vehicles must optimize communications and\nmotion planning jointly. Specifically, a vehicle must adapt its motion plan\nstaying cognizant of communications rate related constraints and adapt the use\nof communications while being cognizant of motion planning related restrictions\nthat may be imposed by the on-road environment. To this end, we formulate a\nreinforcement learning problem wherein an autonomous vehicle jointly chooses\n(a) a motion planning action that executes on-road and (b) a communications\naction of querying sensed information from the infrastructure. The goal is to\noptimize the driving utility of the autonomous vehicle. We apply the Q-learning\nalgorithm to make the vehicle learn the optimal policy, which makes the optimal\nchoice of planning and communications actions at any given time. We demonstrate\nthe ability of the optimal policy to smartly adapt communications and planning\nactions, while achieving large driving utilities, using simulations.\n",
"title": "A Reinforcement Learning Approach to Jointly Adapt Vehicular Communications and Planning for Optimized Driving"
}
| null | null | null | null | true | null |
14124
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the problem of estimating the covariance matrix under\ndifferential privacy, where the underlying covariance matrix is assumed to be\nsparse and of high dimensions. We propose a new method, called DP-Thresholding,\nto achieve a non-trivial $\\ell_2$-norm based error bound, which is\nsignificantly better than the existing ones from adding noise directly to the\nempirical covariance matrix. We also extend the $\\ell_2$-norm based error bound\nto a general $\\ell_w$-norm based one for any $1\\leq w\\leq \\infty$, and show\nthat they share the same upper bound asymptotically. Our approach can be easily\nextended to local differential privacy. Experiments on the synthetic datasets\nshow consistent results with our theoretical claims.\n",
"title": "Differentially Private High Dimensional Sparse Covariance Matrix Estimation"
}
| null | null | null | null | true | null |
14125
| null |
Default
| null | null |
null |
{
"abstract": " Planning safe paths is a major building block in robot autonomy. It has been\nan active field of research for several decades, with a plethora of planning\nmethods. Planners can be generally categorised as either trajectory optimisers\nor sampling-based planners. The latter is the predominant planning paradigm for\noccupancy maps. Trajectory optimisation entails major algorithmic changes to\ntackle contextual information gaps caused by incomplete sensor coverage of the\nmap. However, the benefits are substantial, as trajectory optimisers can reason\non the trade-off between path safety and efficiency.\nIn this work, we improve our previous work on stochastic functional gradient\nplanners. We introduce a novel expressive path representation based on kernel\napproximation, that allows cost effective model updates based on stochastic\nsamples. The main drawback of the previous stochastic functional gradient\nplanner was the cubic cost, stemming from its non-parametric path\nrepresentation. Our novel approximate kernel based model, on the other hand,\nhas a fixed linear cost that depends solely on the number of features used to\nrepresent the path. We show that the stochasticity of the samples is crucial\nfor the planner and present comparisons to other state-of-the-art planning\nmethods in both simulation and with real occupancy data. The experiments\ndemonstrate the advantages of the stochastic approximate kernel method for path\nplanning in occupancy maps.\n",
"title": "Stochastic Functional Gradient Path Planning in Occupancy Maps"
}
| null | null | null | null | true | null |
14126
| null |
Default
| null | null |
null |
{
"abstract": " In display advertising, users' online ad experiences are important for the\nadvertising effectiveness. However, users have not been well accommodated in\nreal-time bidding (RTB). This further influences their site visits and\nperception of the displayed banner ads. In this paper, we propose a novel\ncomputational framework which brings multimedia metrics, like the contextual\nrelevance, the visual saliency and the ad memorability into RTB to improve the\nusers' ad experiences as well as maintain the benefits of the publisher and the\nadvertiser. We aim at developing a vigorous ecosystem by optimizing the\ntrade-offs among all stakeholders. The framework considers the scenario of a\nwebpage with multiple ad slots. Our experimental results show that the benefits\nof the advertiser and the user can be significantly improved if the publisher\nwould slightly sacrifice his short-term revenue. The improved benefits will\nincrease the advertising requests (demand) and the site visits (supply), which\ncan further boost the publisher's revenue in the long run.\n",
"title": "MM2RTB: Bringing Multimedia Metrics to Real-Time Bidding"
}
| null | null | null | null | true | null |
14127
| null |
Default
| null | null |
null |
{
"abstract": " The hallmark of Weyl semimetals is the existence of open constant-energy\ncontours on their surface -- the so-called Fermi arcs -- connecting Weyl\npoints. Here, we show that for time-reversal symmetric realizations of Weyl\nsemimetals these Fermi arcs in many cases coexist with closed Fermi pockets\noriginating from surface Dirac cones pinned to time-reversal invariant momenta.\nThe existence of Fermi pockets is required for certain Fermi-arc connectivities\ndue to additional restrictions imposed by the six $\\mathbb{Z}_2$ topological\ninvariants characterizing a generic time-reversal invariant Weyl semimetal. We\nshow that a change of the Fermi-arc connectivity generally leads to a different\ntopology of the surface Fermi surface, and identify the half-Heusler compound\nLaPtBi under in-plane compressive strain as a material that realizes this\nsurface Lifshitz transition. We also discuss universal features of this\ncoexistence in quasi-particle interference spectra.\n",
"title": "Generic coexistence of Fermi arcs and Dirac cones on the surface of time-reversal invariant Weyl semimetals"
}
| null | null | null | null | true | null |
14128
| null |
Default
| null | null |
null |
{
"abstract": " We study the dynamic response of a superfluid field to a moving edge\ndislocation line to which the field is minimally coupled. We use a dissipative\nGross-Pitaevskii equation, and determine the initial conditions by solving the\nequilibrium version of the model. We consider the subsequent time evolution of\nthe field for both glide and climb dislocation motion and analyze the results\nfor a range of values of the constant speed $V_D$ of the moving dislocation. We\nfind that the type of motion of the dislocation line is very important in\ndetermining the time evolution of the superfluid field distribution associated\nwith it. Climb motion of the dislocation line induces increasing asymmetry, as\nfunction of time, in the field profile, with part of the probability being, as\nit were, left behind. On the other hand, glide motion has no effect on the\nsymmetry properties of the superfluid field distribution. Damping of the\nsuperfluid field due to excitations associated with the moving dislocation line\noccurs in both cases.\n",
"title": "Superfluid Field response to Edge dislocation motion"
}
| null | null |
[
"Physics"
] | null | true | null |
14129
| null |
Validated
| null | null |
null |
{
"abstract": " We examine spectral operator-theoretic properties of linear and nonlinear\ndynamical systems with equilibrium and quasi-periodic attractors and use such\nproperties to characterize a class of datasets and introduce a new notion of\nthe principal dimension of the data.\n",
"title": "Koopman Operator Spectrum and Data Analysis"
}
| null | null | null | null | true | null |
14130
| null |
Default
| null | null |
null |
{
"abstract": " Mobile robots are cyber-physical systems where the cyberspace and the\nphysical world are strongly coupled. Attacks against mobile robots can\ntranscend cyber defenses and escalate into disastrous consequences in the\nphysical world. In this paper, we focus on the detection of active attacks that\nare capable of directly influencing robot mission operation. Through leveraging\nphysical dynamics of mobile robots, we develop RIDS, a novel robot intrusion\ndetection system that can detect actuator attacks as well as sensor attacks for\nnonlinear mobile robots subject to stochastic noises. We implement and evaluate\na RIDS on Khepera mobile robot against concrete attack scenarios via various\nattack channels including signal interference, sensor spoofing, logic bomb, and\nphysical damage. Evaluation of 20 experiments shows that the averages of false\npositive rates and false negative rates are both below 1%. Average detection\ndelay for each attack remains within 0.40s.\n",
"title": "Exploiting Physical Dynamics to Detect Actuator and Sensor Attacks in Mobile Robots"
}
| null | null | null | null | true | null |
14131
| null |
Default
| null | null |
null |
{
"abstract": " In spite of the recent success of neural machine translation (NMT) in\nstandard benchmarks, the lack of large parallel corpora poses a major practical\nproblem for many language pairs. There have been several proposals to alleviate\nthis issue with, for instance, triangulation and semi-supervised learning\ntechniques, but they still require a strong cross-lingual signal. In this work,\nwe completely remove the need of parallel data and propose a novel method to\ntrain an NMT system in a completely unsupervised manner, relying on nothing but\nmonolingual corpora. Our model builds upon the recent work on unsupervised\nembedding mappings, and consists of a slightly modified attentional\nencoder-decoder model that can be trained on monolingual corpora alone using a\ncombination of denoising and backtranslation. Despite the simplicity of the\napproach, our system obtains 15.56 and 10.21 BLEU points in WMT 2014\nFrench-to-English and German-to-English translation. The model can also profit\nfrom small parallel corpora, and attains 21.81 and 15.24 points when combined\nwith 100,000 parallel sentences, respectively. Our implementation is released\nas an open source project.\n",
"title": "Unsupervised Neural Machine Translation"
}
| null | null | null | null | true | null |
14132
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning and data analysis now finds both scientific and industrial\napplication in biology, chemistry, geology, medicine, and physics. These\napplications rely on large quantities of data gathered from automated sensors\nand user input. Furthermore, the dimensionality of many datasets is extreme:\nmore details are being gathered about single user interactions or sensor\nreadings. All of these applications encounter problems with a common theme: use\nobserved data to make inferences about the world. Our work obtains the first\nprovably efficient algorithms for Independent Component Analysis (ICA) in the\npresence of heavy-tailed data. The main tool in this result is the centroid\nbody (a well-known topic in convex geometry), along with optimization and\nrandom walks for sampling from a convex body. This is the first algorithmic use\nof the centroid body and it is of independent theoretical interest, since it\neffectively replaces the estimation of covariance from samples, and is more\ngenerally accessible.\nThis reduction relies on a non-linear transformation of samples from such an\nintersection of halfspaces (i.e. a simplex) to samples which are approximately\nfrom a linearly transformed product distribution. Through this transformation\nof samples, which can be done efficiently, one can then use an ICA algorithm to\nrecover the vertices of the intersection of halfspaces.\nFinally, we again use ICA as an algorithmic primitive to construct an\nefficient solution to the widely-studied problem of learning the parameters of\na Gaussian mixture model. Our algorithm again transforms samples from a\nGaussian mixture model into samples which fit into the ICA model and, when\nprocessed by an ICA algorithm, result in recovery of the mixture parameters.\nOur algorithm is effective even when the number of Gaussians in the mixture\ngrows polynomially with the ambient dimension\n",
"title": "Geometric Methods for Robust Data Analysis in High Dimension"
}
| null | null | null | null | true | null |
14133
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we introduce new modules over the ring of ponderation\nfunctions, so we recover old results in harmonic analysis from the side of ring\ntheory.\nMoreover, we prove that Laplace transform, Fourier transform and Hankel\ntransform generate some kind of modules over the ring of ponderation functions.\n",
"title": "Modules Over the Ring of Ponderation functions with Applications to a Class of Integral Operators"
}
| null | null | null | null | true | null |
14134
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the eigenvalue problem on $\\mathbb{R}^d$ for a class of\nsecond order, elliptic operators of the form $\\mathscr{L} =\na^{ij}\\partial_{x_i}\\partial_{x_j} + b^{i}\\partial_{x_i} + f$, associated with\nnon-degenerate diffusions. We show that strict monotonicity of the principal\neigenvalue of the operator with respect to the potential function $f$ fully\ncharacterizes the ergodic properties of the associated ground state diffusion,\nand the unicity of the ground state, and we present a comprehensive study of\nthe eigenvalue problem from this point of view. This allows us to extend or\nstrengthen various results in the literature for a class of viscous\nHamilton-Jacobi equations of ergodic type with smooth coefficients to equations\nwith measurable drift and potential. In addition, we establish the strong\nduality for the equivalent infinite dimensional linear programming formulation\nof these ergodic control problems. We also apply these results to the study of\nthe infinite horizon risk-sensitive control problem for diffusions, and\nestablish existence of optimal Markov controls, verification of optimality\nresults, and the continuity of the controlled principal eigenvalue with respect\nto stationary Markov controls.\n",
"title": "Strict monotonicity of principal eigenvalues of elliptic operators in $\\mathbb{R}^d$ and risk-sensitive control"
}
| null | null | null | null | true | null |
14135
| null |
Default
| null | null |
null |
{
"abstract": " This report is targeted to groups who are subject matter experts in their\napplication but deep learning novices. It contains practical advice for those\ninterested in testing the use of deep neural networks on applications that are\nnovel for deep learning. We suggest making your project more manageable by\ndividing it into phases. For each phase this report contains numerous\nrecommendations and insights to assist novice practitioners.\n",
"title": "Best Practices for Applying Deep Learning to Novel Applications"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14136
| null |
Validated
| null | null |
null |
{
"abstract": " Building interactive tools to support data analysis is hard because it is not\nalways clear what to build and how to build it. To address this problem, we\npresent Precision Interfaces, a semi-automatic system to generate task-specific\ndata analytics interfaces. Precision Interface can turn a log of executed\nprograms into an interface, by identifying micro-variations between the\nprograms and mapping them to interface components. This paper focuses on SQL\nquery logs, but we can generalize the approach to other languages. Our system\noperates in two steps: it first build an interaction graph, which describes how\nthe queries can be transformed into each other. Then, it finds a set of UI\ncomponents that covers a maximal number of transformations. To restrict the\ndomain of changes to be detected, our system uses a domain-specific language,\nPILang. We give a full description of Precision Interface's components,\nshowcase an early prototype on real program logs and discuss future research\nopportunities.\n",
"title": "Precision Interfaces"
}
| null | null | null | null | true | null |
14137
| null |
Default
| null | null |
null |
{
"abstract": " We consider a classical risk process with arrival of claims following a\nstationary Hawkes process. We study the asymptotic regime when the premium rate\nand the baseline intensity of the claims arrival process are large, and claim\nsize is small. The main goal of this article is to establish a diffusion\napproximation by verifying a functional central limit theorem of this model and\nto compute both the finite-time and infinite-time horizon ruin probabilities.\nNumerical results will also be given.\n",
"title": "Gaussian Approximation of a Risk Model with Stationary Hawkes Arrivals of Claims"
}
| null | null | null | null | true | null |
14138
| null |
Default
| null | null |
null |
{
"abstract": " To conduct a more realistic evaluation on Virtualized Network Functions\nresource allocation algorithms, researches needed data on: (1) potential NFs\nchains (policies), (2) traffic flows passing through these NFs chains, (3) how\nthe dynamic traffic changes affect the NFs (scale out/in) and (4) different\ndata center architectures for the NFC. However, there are no publicly available\nreal data sets on NF chains and traffic that pass through NF chains. Therefore\nwe have used data from previous empirical analyses and made some assumptions to\nderive the required data to evaluate resource allocation algorithms for VNFs.\nWe developed four programs to model the gathered data and generate the required\ndata. All gathered data and data modelling programs are publicly available at\ngithub repository.\n",
"title": "Data Modelling for the Evaluation of Virtualized Network Functions Resource Allocation Algorithms"
}
| null | null | null | null | true | null |
14139
| null |
Default
| null | null |
null |
{
"abstract": " Information systems experience an ever-growing volume of unstructured data,\nparticularly in the form of textual materials. This represents a rich source of\ninformation from which one can create value for people, organizations and\nbusinesses. For instance, recommender systems can benefit from automatically\nunderstanding preferences based on user reviews or social media. However, it is\ndifficult for computer programs to correctly infer meaning from narrative\ncontent. One major challenge is negations that invert the interpretation of\nwords and sentences. As a remedy, this paper proposes a novel learning strategy\nto detect negations: we apply reinforcement learning to find a policy that\nreplicates the human perception of negations based on an exogenous response,\nsuch as a user rating for reviews. Our method yields several benefits, as it\neliminates the former need for expensive and subjective manual labeling in an\nintermediate stage. Moreover, the inferred policy can be used to derive\nstatistical inferences and implications regarding how humans process and act on\nnegations.\n",
"title": "Understanding Negations in Information Processing: Learning from Replicating Human Behavior"
}
| null | null | null | null | true | null |
14140
| null |
Default
| null | null |
null |
{
"abstract": " Recent studies show interest in materials with asymmetric friction forces. We\ninvestigate terminal motion of a solid body with circular contact area. We\nassume that friction forces are asymmetric orthotropic. Two cases of pressure\ndistribution are analyzed: Hertz and Boussinesq laws. Equations for friction\nforce and moment are formulated and solved for these cases. Numer- ical results\nshow significant impact of the asymmetry of friction on the motion. Our results\ncan be used for more accurate prediction of contact behavior of bodies made\nfrom new materials with asymmetric surface textures.\n",
"title": "A study of sliding motion of a solid body on a rough surface with asymmetric friction"
}
| null | null | null | null | true | null |
14141
| null |
Default
| null | null |
null |
{
"abstract": " A common approach to analyzing categorical correlated time series data is to\nfit a generalized linear model (GLM) with past data as covariate inputs. There\nremain challenges to conducting inference for short time series length. By\ntreating the historical data as covariate inputs, standard errors of estimates\nof GLM parameters computed using the empirical Fisher information do not fully\naccount the auto-correlation in the data. To overcome this serious limitation,\nwe derive the exact conditional Fisher information matrix of a general logistic\nautoregressive model with endogenous covariates for any series length $T$.\nMoreover, we also develop an iterative computational formula that allows for\nrelatively easy implementation of the proposed estimator. Our simulation\nstudies show that confidence intervals derived using the exact Fisher\ninformation matrix tend to be narrower than those utilizing the empirical\nFisher information matrix while maintaining type I error rates at or below\nnominal levels. Further, we establish that the exact Fisher information matrix\napproaches, as T tends to infinity, the asymptotic Fisher information matrix\npreviously derived for binary time series data. The developed exact conditional\nFisher information matrix is applied to time-series data on respiratory rate\namong a cohort of expectant mothers where it is found to provide narrower\nconfidence intervals for functionals of scientific interest and lead to greater\nstatistical power when compared to the empirical Fisher information matrix.\n",
"title": "Fisher information matrix of binary time series"
}
| null | null | null | null | true | null |
14142
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study inhomogeneous Diophantine approximation with rational\nnumbers of reduced form. The central object to study is the set $W(f,\\theta)$\nas follows, \\begin{eqnarray*} \\left\\{x\\in [0,1]:\\left\n|x-\\frac{m+\\theta(n)}{n}\\right|<\\frac{f(n)}{n}\\text{ for infinitely many\ncoprime pairs of numbers } m,n\\right\\}, \\end{eqnarray*} where\n$\\{f(n)\\}_{n\\in\\mathbb{N}}$ and $\\{\\theta(n)\\}_{n\\in\\mathbb{N}}$ are sequences\nof real numbers in $[0,1/2]$. We will completely determine the Hausdorff\ndimension of $W(f,\\theta)$ in terms of $f$ and $\\theta$. As a by-product, we\nalso obtain a new sufficient condition for $W(f,\\theta)$ to have full Lebesgue\nmeasure and this result is closely related to the study of \\ds with extra\nconditions.\n",
"title": "A Fourier analytic approach to inhomogeneous Diophantine approximation"
}
| null | null | null | null | true | null |
14143
| null |
Default
| null | null |
null |
{
"abstract": " A specific value for the cosmological constant, \\Lambda, can account for\nlate-time cosmic acceleration. However, motivated by the so-called cosmological\nconstant problem(s), several alternative mechanisms have been explored. To\ndate, a host of well-studied dynamical dark energy and modified gravity models\nexists. Going beyond \\Lambda CDM often comes with additional degrees of freedom\n(dofs). For these to pass existing observational tests, an efficient screening\nmechanism must be in place. The linear and quasi-linear regimes of structure\nformation are ideal probes of such dofs and can capture the onset of screening.\nWe propose here a semi-phenomenological treatment to account for screening\ndynamics on LSS observables, with special emphasis on Vainshtein-type\nscreening.\n",
"title": "Screening in perturbative approaches to LSS"
}
| null | null | null | null | true | null |
14144
| null |
Default
| null | null |
null |
{
"abstract": " Synthesis of rationally designed nanostructured materials with optimized\nmechanical properties, e.g., high strength with considerable ductility,\nrequires rigorous control of diverse microstructural parameters including the\nmean size, size dispersion and spatial distribution of grains. However,\ncurrently available synthesis techniques can seldom satisfy these requirements.\nHere, we report a new methodology to synthesize thin films with unprecedented\nmicrostructural control via systematic, in-situ seeding of nanocrystals into\namorphous precursor films. When the amorphous films are subsequently\ncrystallized by thermal annealing, the nanocrystals serve as preferential grain\nnucleation sites and control their microstructure. We demonstrate the\ncapability of this approach by precisely tailoring the size, geometry and\nspatial distribution of nanostructured grains in structural (TiAl) as well as\nfunctional (TiNi) thin films. The approach, which is applicable to a broad\nclass of metallic alloys and ceramics, enables explicit microstructural control\nof thin film materials for a wide spectrum of applications.\n",
"title": "Thin films with precisely engineered nanostructures"
}
| null | null |
[
"Physics"
] | null | true | null |
14145
| null |
Validated
| null | null |
null |
{
"abstract": " We propose a novel framework that reduces the management and integration\noverheads for real-time network flows by leveraging the capabilities\n(especially global visibility and management) of software-defined networking\n(SDN) architectures. Given the specifications of flows that must meet hard\nreal-time requirements, our framework synthesizes paths through the network and\nassociated switch configurations - to guarantee that these flows meet their\nend-to-end timing requirements. In doing so, our framework makes SDN\narchitectures \"delay-aware\" - remember that SDN is otherwise not able to reason\nabout delays. Hence, it is easier to use such architectures in safety-critical\nand other latency-sensitive applications. We demonstrate our principles as well\nas the feasibility of our approach using both - exhaustive simulations as well\nas experiments using real hardware switches.\n",
"title": "End-to-End Network Delay Guarantees for Real-Time Systems using SDN"
}
| null | null | null | null | true | null |
14146
| null |
Default
| null | null |
null |
{
"abstract": " We present semi-analytical models of galactic outflows in high redshift\ngalaxies driven by both hot thermal gas and non-thermal cosmic rays. Thermal\npressure alone may not sustain a large scale outflow in low mass galaxies (i.e\n$M\\sim 10^8$~M$_\\odot$), in the presence of supernovae (SNe) feedback with\nlarge mass loading. We show that inclusion of cosmic ray pressure allows\noutflow solutions even in these galaxies. In massive galaxies for the same\nenergy efficiency, cosmic ray driven winds can propagate to larger distances\ncompared to pure thermally driven winds. On an average gas in the cosmic ray\ndriven winds has a lower temperature which could aid detecting it through\nabsorption lines in the spectra of background sources. Using our constrained\nsemi-analytical models of galaxy formation (that explains the observed UV\nluminosity functions of galaxies) we study the influence of cosmic ray driven\nwinds on the properties of the intergalactic medium (IGM) at different\nredshifts. In particular, we study the volume filling factor, average\nmetallicity, cosmic ray and magnetic field energy densities for models invoking\natomic cooled and molecular cooled halos. We show that the cosmic rays in the\nIGM could have enough energy that can be transferred to the thermal gas in\npresence of magnetic fields to influence the thermal history of the\nintergalactic medium. The significant volume filling and resulting strength of\nIGM magnetic fields can also account for recent $\\gamma$-ray observations of\nblazars.\n",
"title": "Efficient cold outflows driven by cosmic rays in high redshift galaxies and their global effects on the IGM"
}
| null | null | null | null | true | null |
14147
| null |
Default
| null | null |
null |
{
"abstract": " Learning algorithms that learn linear models often have high representation\nbias on real-world problems. In this paper, we show that this representation\nbias can be greatly reduced by discretization. Discretization is a common\nprocedure in machine learning that is used to convert a quantitative attribute\ninto a qualitative one. It is often motivated by the limitation of some\nlearners to qualitative data. Discretization loses information, as fewer\ndistinctions between instances are possible using discretized data relative to\nundiscretized data. In consequence, where discretization is not essential, it\nmight appear desirable to avoid it. However, it has been shown that\ndiscretization often substantially reduces the error of the linear generative\nBayesian classifier naive Bayes. This motivates a systematic study of the\neffectiveness of discretizing quantitative attributes for other linear\nclassifiers. In this work, we study the effect of discretization on the\nperformance of linear classifiers optimizing three distinct discriminative\nobjective functions --- logistic regression (optimizing negative\nlog-likelihood), support vector classifiers (optimizing hinge loss) and a\nzero-hidden layer artificial neural network (optimizing mean-square-error). We\nshow that discretization can greatly increase the accuracy of these linear\ndiscriminative learners by reducing their representation bias, especially on\nbig datasets. We substantiate our claims with an empirical study on $42$\nbenchmark datasets.\n",
"title": "On the Effectiveness of Discretizing Quantitative Attributes in Linear Classifiers"
}
| null | null |
[
"Computer Science"
] | null | true | null |
14148
| null |
Validated
| null | null |
null |
{
"abstract": " Kustaanheimo-Stiefel (KS) transformation depends on the choice of some\npreferred direction in the Cartesian 3D space. This choice, seldom explicitly\nmentioned, amounts typically to the direction of the first or the third\ncoordinate axis in celestial mechanics and atomic physics, respectively. The\npresent work develops a canonical KS transformation with an arbitrary preferred\ndirection, indicated by what we call a defining vector. Using a mix of vector\nand quaternion algebra, we formulate the transformation in a reference frame\nindependent manner. The link between the oscillator and Keplerian first\nintegrals is given. As an example of the present formulation, the Keplerian\nmotion in a rotating frame is re-investigated.\n",
"title": "Kustaanheimo-Stiefel transformation with an arbitrary defining vector"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14149
| null |
Validated
| null | null |
null |
{
"abstract": " This paper presents a new safety specification method that is robust against\nerrors in the probability distribution of disturbances. Our proposed\ndistributionally robust safe policy maximizes the probability of a system\nremaining in a desired set for all times, subject to the worst possible\ndisturbance distribution in an ambiguity set. We propose a dynamic game\nformulation of constructing such policies and identify conditions under which a\nnon-randomized Markov policy is optimal. Based on this existence result, we\ndevelop a practical design approach to safety-oriented stochastic controllers\nwith limited information about disturbance distributions. This control method\ncan be used to minimize another cost function while ensuring safety in a\nprobabilistic way. However, an associated Bellman equation involves\ninfinite-dimensional minimax optimization problems since the disturbance\ndistribution may have a continuous density. To resolve computational issues, we\npropose a duality-based reformulation method that converts the\ninfinite-dimensional minimax problem into a semi-infinite program that can be\nsolved using existing convergent algorithms. We prove that there is no duality\ngap, and that this approach thus preserves optimality. The results of numerical\ntests confirm that the proposed method is robust against distributional errors\nin disturbances, while a standard stochastic safety specification tool is not.\n",
"title": "A dynamic game approach to distributionally robust safety specifications for stochastic systems"
}
| null | null | null | null | true | null |
14150
| null |
Default
| null | null |
null |
{
"abstract": " We present a simple yet effective approach for linking entities in queries.\nThe key idea is to search sentences similar to a query from Wikipedia articles\nand directly use the human-annotated entities in the similar sentences as\ncandidate entities for the query. Then, we employ a rich set of features, such\nas link-probability, context-matching, word embeddings, and relatedness among\ncandidate entities as well as their related entities, to rank the candidates\nunder a regression based framework. The advantages of our approach lie in two\naspects, which contribute to the ranking process and final linking result.\nFirst, it can greatly reduce the number of candidate entities by filtering out\nirrelevant entities with the words in the query. Second, we can obtain the\nquery sensitive prior probability in addition to the static link-probability\nderived from all Wikipedia articles. We conduct experiments on two benchmark\ndatasets on entity linking for queries, namely the ERD14 dataset and the GERDAQ\ndataset. Experimental results show that our method outperforms state-of-the-art\nsystems and yields 75.0% in F1 on the ERD14 dataset and 56.9% on the GERDAQ\ndataset.\n",
"title": "Entity Linking for Queries by Searching Wikipedia Sentences"
}
| null | null | null | null | true | null |
14151
| null |
Default
| null | null |
null |
{
"abstract": " Reinforcement Learning AI commonly uses reward/penalty signals that are\nobjective and explicit in an environment -- e.g. game score, completion time,\netc. -- in order to learn the optimal strategy for task performance. However,\nHuman-AI interaction for such AI agents should include additional reinforcement\nthat is implicit and subjective -- e.g. human preferences for certain AI\nbehavior -- in order to adapt the AI behavior to idiosyncratic human\npreferences. Such adaptations would mirror naturally occurring processes that\nincrease trust and comfort during social interactions. Here, we show how a\nhybrid brain-computer-interface (hBCI), which detects an individual's level of\ninterest in objects/events in a virtual environment, can be used to adapt the\nbehavior of a Deep Reinforcement Learning AI agent that is controlling a\nvirtual autonomous vehicle. Specifically, we show that the AI learns a driving\nstrategy that maintains a safe distance from a lead vehicle, and most novelly,\npreferentially slows the vehicle when the human passengers of the vehicle\nencounter objects of interest. This adaptation affords an additional 20\\%\nviewing time for subjectively interesting objects. This is the first\ndemonstration of how an hBCI can be used to provide implicit reinforcement to\nan AI agent in a way that incorporates user preferences into the control\nsystem.\n",
"title": "Towards personalized human AI interaction - adapting the behavior of AI agents using neural signatures of subjective interest"
}
| null | null | null | null | true | null |
14152
| null |
Default
| null | null |
null |
{
"abstract": " Memory-based neural networks model temporal data by leveraging an ability to\nremember information for long periods. It is unclear, however, whether they\nalso have an ability to perform complex relational reasoning with the\ninformation they remember. Here, we first confirm our intuitions that standard\nmemory architectures may struggle at tasks that heavily involve an\nunderstanding of the ways in which entities are connected -- i.e., tasks\ninvolving relational reasoning. We then improve upon these deficits by using a\nnew memory module -- a \\textit{Relational Memory Core} (RMC) -- which employs\nmulti-head dot product attention to allow memories to interact. Finally, we\ntest the RMC on a suite of tasks that may profit from more capable relational\nreasoning across sequential information, and show large gains in RL domains\n(e.g. Mini PacMan), program evaluation, and language modeling, achieving\nstate-of-the-art results on the WikiText-103, Project Gutenberg, and GigaWord\ndatasets.\n",
"title": "Relational recurrent neural networks"
}
| null | null | null | null | true | null |
14153
| null |
Default
| null | null |
null |
{
"abstract": " There is currently great interest in applying neural networks to prediction\ntasks in medicine. It is important for predictive models to be able to use\nsurvival data, where each patient has a known follow-up time and\nevent/censoring indicator. This avoids information loss when training the model\nand enables generation of predicted survival curves. In this paper, we describe\na discrete-time survival model that is designed to be used with neural\nnetworks, which we refer to as Nnet-survival. The model is trained with the\nmaximum likelihood method using minibatch stochastic gradient descent (SGD).\nThe use of SGD enables rapid convergence and application to large datasets that\ndo not fit in memory. The model is flexible, so that the baseline hazard rate\nand the effect of the input data on hazard probability can vary with follow-up\ntime. It has been implemented in the Keras deep learning framework, and source\ncode for the model and several examples is available online. We demonstrate the\nperformance of the model on both simulated and real data and compare it to\nexisting models Cox-nnet and Deepsurv.\n",
"title": "A Scalable Discrete-Time Survival Model for Neural Networks"
}
| null | null | null | null | true | null |
14154
| null |
Default
| null | null |
null |
{
"abstract": " An infinite chain of driven-dissipative condensate spins with uniform\nnearest-neighbor coherent coupling is solved analytically and investigated\nnumerically. Above a critical occupation threshold the condensates undergo\nspontaneous spin bifurcation (becoming magnetized) forming a binary chain of\nspin-up or spin-down states. Minimization of the bifurcation threshold\ndetermines the magnetic order as a function of the coupling strength. This\nallows control of multiple magnetic orders via adiabatic (slow ramping of)\npumping. In addition to ferromagnetic and anti-ferromagnetic ordered states we\nshow the formation of a paired-spin ordered state $\\left|\\dots \\uparrow\n\\uparrow \\downarrow \\downarrow \\dots \\right. \\rangle$ as a consequence of the\nphase degree of freedom between condensates.\n",
"title": "A driven-dissipative spin chain model based on exciton-polariton condensates"
}
| null | null |
[
"Physics"
] | null | true | null |
14155
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, a projected primal-dual gradient flow of augmented Lagrangian\nis presented to solve convex optimization problems that are not necessarily\nstrictly convex. The optimization variables are restricted by a convex set with\ncomputable projection operation on its tangent cone as well as equality\nconstraints. As a supplement of the analysis in\n\\cite{niederlander2016distributed}, we show that the projected dynamical system\nconverges to one of the saddle points and hence finding an optimal solution.\nMoreover, the problem of distributedly maximizing the algebraic connectivity of\nan undirected network by optimizing the port gains of each nodes (base\nstations) is considered. The original semi-definite programming (SDP) problem\nis relaxed into a nonlinear programming (NP) problem that will be solved by the\naforementioned projected dynamical system. Numerical examples show the\nconvergence of the aforementioned algorithm to one of the optimal solutions.\nThe effect of the relaxation is illustrated empirically with numerical\nexamples. A methodology is presented so that the number of iterations needed to\nreach the equilibrium is suppressed. Complexity per iteration of the algorithm\nis illustrated with numerical examples.\n",
"title": "Projected Primal-Dual Gradient Flow of Augmented Lagrangian with Application to Distributed Maximization of the Algebraic Connectivity of a Network"
}
| null | null | null | null | true | null |
14156
| null |
Default
| null | null |
null |
{
"abstract": " Although there is a significant literature on the asymptotic theory of Bayes\nfactor, the set-ups considered are usually specialized and often involves\nindependent and identically distributed data. Even in such specialized cases,\nmostly weak consistency results are available. In this article, for the first\ntime ever, we derive the almost sure convergence theory of Bayes factor in the\ngeneral set-up that includes even dependent data and misspecified models.\nSomewhat surprisingly, the key to the proof of such a general theory is a\nsimple application of a result of Shalizi (2009) to a well-known identity\nsatisfied by the Bayes factor.\n",
"title": "A Short Note on Almost Sure Convergence of Bayes Factors in the General Set-Up"
}
| null | null | null | null | true | null |
14157
| null |
Default
| null | null |
null |
{
"abstract": " Aharoni and Berger conjectured that in any bipartite multigraph that is\nproperly edge-coloured by $n$ colours with at least $n + 1$ edges of each\ncolour there must be a matching that uses each colour exactly once. In this\npaper we consider the same question without the bipartiteness assumption. We\nshow that in any multigraph with edge multiplicities $o(n)$ that is properly\nedge-coloured by $n$ colours with at least $n + o(n)$ edges of each colour\nthere must be a matching of size $n-O(1)$ that uses each colour at most once.\n",
"title": "Rainbow matchings in properly-coloured multigraphs"
}
| null | null | null | null | true | null |
14158
| null |
Default
| null | null |
null |
{
"abstract": " This paper provides a general and abstract approach to approximate ergodic\nregimes of Markov and Feller processes. More precisely, we show that the\nrecursive algorithm presented in Lamberton & Pages (2002) and based on\nsimulation algorithms of stochastic schemes with decreasing step can be used to\nbuild invariant measures for general Markov and Feller processes. We also\npropose applications in three different configurations: Approximation of Markov\nswitching Brownian diffusion ergodic regimes using Euler scheme, approximation\nof Markov Brownian diffusion ergodic regimes with Milstein scheme and\napproximation of general diffusions with jump components ergodic regimes.\n",
"title": "Recursive computation of the invariant distribution of Markov and Feller processes"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14159
| null |
Validated
| null | null |
null |
{
"abstract": " One of the most interesting features in the libration domain of co-orbital\nmotions is the existence of secondary resonances. For some combinations of\nphysical parameters, these resonances occupy a large fraction of the domain of\nstability and rule the dynamics within the stable tadpole region. In this work,\nwe present an application of a recently introduced `basic Hamiltonian model' Hb\nfor Trojan dynamics, in Paez and Efthymiopoulos (2015), Paez, Locatelli and\nEfthymiopoulos (2016): we show that the inner border of the secondary resonance\nof lowermost order, as defined by Hb, provides a good estimation of the region\nin phase-space for which the orbits remain regular regardless the orbital\nparameters of the system. The computation of this boundary is straightforward\nby combining a resonant normal form calculation in conjunction with an\n`asymmetric expansion' of the Hamiltonian around the libration points, which\nspeeds up convergence. Applications to the determination of the effective\nstability domain for exoplanetary Trojans (planet-sized objects or asteroids)\nwhich may accompany giant exoplanets are discussed.\n",
"title": "Secondary resonances and the boundary of effective stability of Trojan motions"
}
| null | null | null | null | true | null |
14160
| null |
Default
| null | null |
null |
{
"abstract": " The deep Q-network (DQN) and return-based reinforcement learning are two\npromising algorithms proposed in recent years. DQN brings advances to complex\nsequential decision problems, while return-based algorithms have advantages in\nmaking use of sample trajectories. In this paper, we propose a general\nframework to combine DQN and most of the return-based reinforcement learning\nalgorithms, named R-DQN. We show the performance of traditional DQN can be\nimproved effectively by introducing return-based reinforcement learning. In\norder to further improve the R-DQN, we design a strategy with two measurements\nwhich can qualitatively measure the policy discrepancy. Moreover, we give the\ntwo measurements' bounds in the proposed R-DQN framework. We show that\nalgorithms with our strategy can accurately express the trace coefficient and\nachieve a better approximation to return. The experiments, conducted on several\nrepresentative tasks from the OpenAI Gym library, validate the effectiveness of\nthe proposed measurements. The results also show that the algorithms with our\nstrategy outperform the state-of-the-art methods.\n",
"title": "Qualitative Measurements of Policy Discrepancy for Return-based Deep Q-Network"
}
| null | null | null | null | true | null |
14161
| null |
Default
| null | null |
null |
{
"abstract": " Recent studies regarding the habitability, observability, and possible\norbital evolution of the indirectly detected exoplanet Proxima b have mostly\nassumed a planet with $M \\sim 1.3$ $M_\\oplus$, a rocky composition, and an\nEarth-like atmosphere or none at all. In order to assess these assumptions, we\nuse previous studies of the radii, masses, and compositions of super-Earth\nexoplanets to probabilistically constrain the mass and radius of Proxima b,\nassuming an isotropic inclination probability distribution. We find it is ~90%\nlikely that the planet's density is consistent with a rocky composition;\nconversely, it is at least 10% likely that the planet has a significant amount\nof ice or an H/He envelope. If the planet does have a rocky composition, then\nwe find expectation values and 95% confidence intervals of\n$\\left<M\\right>_\\text{rocky} = 1.63_{-0.72}^{+1.66}$ $M_\\oplus$ for its mass\nand $\\left<R\\right>_\\text{rocky} = 1.07_{-0.31}^{+0.38}$ $R_\\oplus$ for its\nradius.\n",
"title": "Probabilistic Constraints on the Mass and Composition of Proxima b"
}
| null | null | null | null | true | null |
14162
| null |
Default
| null | null |
null |
{
"abstract": " We exhibit the first explicit examples of Salem sets in $\\mathbb{Q}_p$ of\nevery dimension $0 < \\alpha < 1$ by showing that certain sets of\nwell-approximable $p$-adic numbers are Salem sets. We construct measures\nsupported on these sets that satisfy essentially optimal Fourier decay and\nupper regularity conditions, and we observe that these conditions imply that\nthe measures satisfy strong Fourier restriction inequalities. We also partially\ngeneralize our results to higher dimensions. Our results extend theorems of\nKaufman, Papadimitropoulos, and Hambrook from the real to the $p$-adic setting.\n",
"title": "Explicit Salem sets, Fourier restriction, and metric Diophantine approximation in the $p$-adic numbers"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14163
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the hard-core model on finite triangular lattices with Metropolis\ndynamics. Under suitable conditions on the triangular lattice dimensions, this\ninteracting particle system has three maximum-occupancy configurations and we\ninvestigate its high-fugacity behavior by studying tunneling times, i.e., the\nfirst hitting times between between these maximum-occupancy configurations, and\nthe mixing time. The proof method relies on the analysis of the corresponding\nstate space using geometrical and combinatorial properties of the hard-core\nconfigurations on finite triangular lattices, in combination with known results\nfor first hitting times of Metropolis Markov chains in the equivalent\nzero-temperature limit. In particular, we show how the order of magnitude of\nthe expected tunneling times depends on the triangular lattice dimensions in\nthe low-temperature regime and prove the asymptotic exponentiality of the\nrescaled tunneling time leveraging the intrinsic symmetry of the state space.\n",
"title": "Tunneling of the hard-core model on finite triangular lattices"
}
| null | null | null | null | true | null |
14164
| null |
Default
| null | null |
null |
{
"abstract": " Model reduction of the Markov process is a basic problem in modeling\nstate-transition systems. Motivated by the state aggregation approach rooted in\ncontrol theory, we study the statistical state compression of a finite-state\nMarkov chain from empirical trajectories. Through the lens of spectral\ndecomposition, we study the rank and features of Markov processes, as well as\nproperties like representability, aggregatability, and lumpability. We develop\na class of spectral state compression methods for three tasks: (1) estimate the\ntransition matrix of a low-rank Markov model, (2) estimate the leading subspace\nspanned by Markov features, and (3) recover latent structures of the state\nspace like state aggregation and lumpable partition. The proposed methods\nprovide an unsupervised learning framework for identifying Markov features and\nclustering states. We provide upper bounds for the estimation errors and nearly\nmatching minimax lower bounds. Numerical studies are performed on synthetic\ndata and a dataset of New York City taxi trips.\n",
"title": "Spectral State Compression of Markov Processes"
}
| null | null | null | null | true | null |
14165
| null |
Default
| null | null |
null |
{
"abstract": " We have carried out a campaign to characterize the hot Jupiters WASP-5b,\nWASP-44b and WASP-46b using multiband photometry collected at the\nObservatório do Pico Dos Dias in Brazil. We have determined the planetary\nphysical properties and new transit ephemerides for these systems. The new\norbital parameters and physical properties of WASP-5b and WASP-44b are\nconsistent with previous estimates. In the case of WASP-46b, there is some\nquota of disagreement between previous results. We provide a new determination\nof the radius of this planet and help clarify the previous differences. We also\nstudied the transit time variations including our new measurements. No clear\nvariation from a linear trend was found for the systems WASP-5b and WASP-44b.\nIn the case of WASP-46b, we found evidence of deviations indicating the\npresence of a companion but statistical analysis of the existing times points\nto a signal due to the sampling rather than a new planet. Finally, we studied\nthe fractional radius variation as a function of wavelength for these systems.\nThe broad-band spectrums of WASP-5b and WASP-44b are mostly flat. In the case\nof WASP-46b we found a trend, but further measurements are necessary to confirm\nthis finding.\n",
"title": "Multi-band characterization of the hot Jupiters: WASP-5b, WASP-44b and WASP-46b"
}
| null | null | null | null | true | null |
14166
| null |
Default
| null | null |
null |
{
"abstract": " When a 2D superconductor is subjected to a strong in-plane magnetic field,\nZeeman polarization of the Fermi surface can give rise to inhomogeneous FFLO\norder with a spatially modulated gap. Further increase of the magnetic field\neventually drives the system into a normal metal state. Here, we perform a\nrenormalization group analysis of this quantum phase transition, starting from\nan appropriate low-energy theory recently introduced by Piazza et al. (Ref.1).\nWe compute one-loop flow equations within the controlled dimensional\nregularization scheme with fixed dimension of Fermi surface, expanding in\n$\\epsilon = 5/2 - d$. We find a new stable non-Fermi liquid fixed point and\ndiscuss its critical properties. One of the most interesting aspects of the\nFFLO non-Fermi liquid scenario is that the quantum critical point is\npotentially naked, with the scaling regime observable down to arbitrary low\ntemperatures. In order to study this possibility, we perform a general analysis\nof competing instabilities, which suggests that only charge density wave order\nis enhanced in the vicinity of the quantum critical point.\n",
"title": "Non-Fermi liquid at the FFLO quantum critical point"
}
| null | null |
[
"Physics"
] | null | true | null |
14167
| null |
Validated
| null | null |
null |
{
"abstract": " Consider a regression problem where there is no labeled data and the only\nobservations are the predictions $f_i(x_j)$ of $m$ experts $f_{i}$ over many\nsamples $x_j$. With no knowledge on the accuracy of the experts, is it still\npossible to accurately estimate the unknown responses $y_{j}$? Can one still\ndetect the least or most accurate experts? In this work we propose a framework\nto study these questions, based on the assumption that the $m$ experts have\nuncorrelated deviations from the optimal predictor. Assuming the first two\nmoments of the response are known, we develop methods to detect the best and\nworst regressors, and derive U-PCR, a novel principal components approach for\nunsupervised ensemble regression. We provide theoretical support for U-PCR and\nillustrate its improved accuracy over the ensemble mean and median on a variety\nof regression problems.\n",
"title": "Unsupervised Ensemble Regression"
}
| null | null | null | null | true | null |
14168
| null |
Default
| null | null |
null |
{
"abstract": " Although Poisson-Voronoi diagrams have interesting mathematical properties,\nthere is still much to discover about the geometrical properties of its grains.\nThrough simulations, many authors were able to obtain numerical approximations\nof the moments of the distributions of more or less all geometrical\ncharacteristics of the grain. Furthermore, many proposals on how to get close\nparametric approximations to the real distributions were put forward by several\nauthors. In this paper we show that exploiting the scaling property of the\nunderlying Poisson process, we are able to derive the distribution of the main\ngeometrical features of the grain for every value of the intensity parameter.\nMoreover, we use a sophisticated simulation program to construct a close Monte\nCarlo based approximation for the distributions of interest. Using this, we\nalso determine the closest approximating distributions within the mentioned\nfrequently used parametric classes of distributions and conclude that these\napproximations can be quite accurate.\n",
"title": "Accurate approximation of the distributions of the 3D Poisson-Voronoi typical cell geometrical features"
}
| null | null | null | null | true | null |
14169
| null |
Default
| null | null |
null |
{
"abstract": " One promising avenue to study one-dimensional ($1$D) topological phases is to\nrealize them in synthetic materials such as cold atomic gases. Intriguingly, it\nis possible to realize Majorana boundary modes in a $1$D number-conserving\nsystem consisting of two fermionic chains coupled only by pair-hopping\nprocesses. It is commonly believed that significant interchain single-particle\ntunneling necessarily destroys these Majorana modes, as it spoils the\n$\\mathbb{Z}_2$ fermion parity symmetry that protects them. In this Letter, we\npresent a new mechanism to overcome this obstacle, by piercing a (synthetic)\nmagnetic $\\pi$-flux through each plaquette of the Fermi ladder. Using\nbosonization, we show that in this case there exists an exact leg-interchange\nsymmetry that is robust to interchain hopping, and acts as fermion parity at\nlong wavelengths. We utilize density matrix renormalization group and exact\ndiagonalization to verify that the resulting model exhibits Majorana boundary\nmodes up to large single-particle tunnelings, comparable to the intrachain\nhopping strength. Our work highlights the unusual impacts of different\ntopologically trivial band structures on these interaction-driven topological\nphases, and identifies a distinct route to stabilizing Majorana boundary modes\nin $1$D fermionic ladders.\n",
"title": "Flux-Stabilized Majorana Zero Modes in Coupled One-Dimensional Fermi Wires"
}
| null | null | null | null | true | null |
14170
| null |
Default
| null | null |
null |
{
"abstract": " The use of Kalman filtering, as well as its nonlinear extensions, for the\nestimation of system variables and parameters has played a pivotal role in many\nfields of scientific inquiry where observations of the system are restricted to\na subset of variables. However in the case of censored observations, where\nmeasurements of the system beyond a certain detection point are impossible, the\nestimation problem is complicated. Without appropriate consideration, censored\nobservations can lead to inaccurate estimates. Motivated by the work of [1], we\ndevelop a modified version of the extended Kalman filter to handle the case of\ncensored observations in nonlinear systems. We validate this methodology in a\nsimple oscillator system first, showing its ability to accurately reconstruct\nstate variables and track system parameters when observations are censored.\nFinally, we utilize the nonlinear censored filter to analyze censored datasets\nfrom patients with hepatitis C and human immunodeficiency virus.\n",
"title": "Nonlinear Kalman Filtering for Censored Observations"
}
| null | null | null | null | true | null |
14171
| null |
Default
| null | null |
null |
{
"abstract": " Advances in atomic resolution in situ environmental transmission electron\nmicroscopy for direct probing of gas-solid reactions, including at very high\ntemperatures are described. In addition, recent developments of dynamic real\ntime in situ studies at the Angstrom level using a hot stage in an aberration\ncorrected environment are presented. In situ data from Pt and Pd nanoparticles\non carbon with the corresponding FFT (optical diffractogram) illustrate an\nachieved resolution of 0.11 nm at 500 C and higher in a double aberration\ncorrected TEM and STEM instrument employing a wider gap objective pole piece.\nThe new results open up opportunities for dynamic studies of materials in an\naberration corrected environment.\n",
"title": "Advances in Atomic Resolution In Situ Environmental Transmission Electron Microscopy and 1 Angstrom Aberration Corrected In Situ Electron Microscopy"
}
| null | null | null | null | true | null |
14172
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents an extension of a recently developed high order finite\ndifference method for the wave equation on a grid with non-conforming\ninterfaces. The stability proof of the existing methods relies on the\ninterpolation operators being norm-contracting, which is satisfied by the\nsecond and fourth order operators, but not by the sixth order operator. We\nconstruct new penalty terms to impose interface conditions such that the\nstability proof does not require the norm-contracting condition. As a\nconsequence, the sixth order accurate scheme is also provably stable. Numerical\nexperiments demonstrate the improved stability and accuracy property.\n",
"title": "An improved high order finite difference method for non-conforming grid interfaces for the wave equation"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14173
| null |
Validated
| null | null |
null |
{
"abstract": " In their work on a sharp compactness theorem for the Yamabe problem, Khuri,\nMarques and Schoen apply a refined blow-up analysis (what we call `second order\nblow-up argument' in this article) to obtain highly accurate approximate\nsolutions for the Yamabe equation. As for the conformal scalar curvature\nequation on S^n with n > 3, we examine the second order blow-up argument and\nobtain refined estimate for a blow-up sequence near a simple blow-up point. The\nestimate involves local effect from the Taylor expansion of the scalar\ncurvature function, global effect from other blow-up points, and the balance\nformula as expressed in the Pohozaev identity in an essential way.\n",
"title": "Refined estimates for simple blow-ups of the scalar curvature equation on S^n"
}
| null | null | null | null | true | null |
14174
| null |
Default
| null | null |
null |
{
"abstract": " In this brief note we connect the discrete logarithm problem over prime\nfields in the safe prime case to the logarithmic derivative.\n",
"title": "The discrete logarithm problem over prime fields: the safe prime case. The Smart attack, non-canonical lifts and logarithmic derivatives"
}
| null | null | null | null | true | null |
14175
| null |
Default
| null | null |
null |
{
"abstract": " Convolutional neural networks (CNNs) can be applied to graph similarity\nmatching, in which case they are called graph CNNs. Graph CNNs are attracting\nincreasing attention due to their effectiveness and efficiency. However, the\nexisting convolution approaches focus only on regular data forms and require\nthe transfer of the graph or key node neighborhoods of the graph into the same\nfixed form. During this transfer process, structural information of the graph\ncan be lost, and some redundant information can be incorporated. To overcome\nthis problem, we propose the disordered graph convolutional neural network\n(DGCNN) based on the mixed Gaussian model, which extends the CNN by adding a\npreprocessing layer called the disordered graph convolutional layer (DGCL). The\nDGCL uses a mixed Gaussian function to realize the mapping between the\nconvolution kernel and the nodes in the neighborhood of the graph. The output\nof the DGCL is the input of the CNN. We further implement a\nbackward-propagation optimization process of the convolutional layer by which\nwe incorporate the feature-learning model of the irregular node neighborhood\nstructure into the network. Thereafter, the optimization of the convolution\nkernel becomes part of the neural network learning process. The DGCNN can\naccept arbitrary scaled and disordered neighborhood graph structures as the\nreceptive fields of CNNs, which reduces information loss during graph\ntransformation. Finally, we perform experiments on multiple standard graph\ndatasets. The results show that the proposed method outperforms the\nstate-of-the-art methods in graph classification and retrieval.\n",
"title": "DGCNN: Disordered Graph Convolutional Neural Network Based on the Gaussian Mixture Model"
}
| null | null | null | null | true | null |
14176
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies a \\textit{partial functional partially linear single-index\nmodel} that consists of a functional linear component as well as a linear\nsingle-index component. This model generalizes many well-known existing models\nand is suitable for more complicated data structures. However, its estimation\ninherits the difficulties and complexities from both components and makes it a\nchallenging problem, which calls for new methodology. We propose a novel\nprofile B-spline method to estimate the parameters by approximating the unknown\nnonparametric link function in the single-index component part with B-spline,\nwhile the linear slope function in the functional component part is estimated\nby the functional principal component basis. The consistency and asymptotic\nnormality of the parametric estimators are derived, and the global convergence\nof the proposed estimator of the linear slope function is also established.\nMore excitingly, the latter convergence is optimal in the minimax sense. A\ntwo-stage procedure is implemented to estimate the nonparametric link function,\nand the resulting estimator possesses the optimal global rate of convergence.\nFurthermore, the convergence rate of the mean squared prediction error for a\npredictor is also obtained. Empirical properties of the proposed procedures are\nstudied through Monte Carlo simulations. A real data example is also analyzed\nto illustrate the power and flexibility of the proposed methodology.\n",
"title": "Profile Estimation for Partial Functional Partially Linear Single-Index Model"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
14177
| null |
Validated
| null | null |
null |
{
"abstract": " Wireless rechargeable sensor networks, consisting of sensor nodes with\nrechargeable batteries and mobile chargers to replenish their batteries, have\ngradually become a promising solution to the bottleneck of energy limitation\nthat hinders the wide deployment of wireless sensor networks (WSN). In this\npaper, we focus on the mobile charger scheduling and path optimization scenario\nin which the $k$-coverage ability of a network system needs to be maintained.\nWe formulate the optimal $k$-coverage charging problem of finding a feasible\npath for a mobile charger to charge a set of sensor nodes within their\nestimated charging time windows under the constraint of maintaining the\n$k$-coverage ability of the network system, with an objective of minimizing the\nenergy consumption on traveling per tour. We show the hardness of the problem\nthat even finding a feasible path for the trivial case of the problem is an\nNP-complete one with no polytime constant-factor approximation algorithm.\n",
"title": "Optimal $k$-Coverage Charging Problem"
}
| null | null | null | null | true | null |
14178
| null |
Default
| null | null |
null |
{
"abstract": " Online social media are information resources that can have a transformative\npower in society. While the Web was envisioned as an equalizing force that\nallows everyone to access information, the digital divide prevents large\namounts of people from being present online. Online social media in particular\nare prone to gender inequality, an important issue given the link between\nsocial media use and employment. Understanding gender inequality in social\nmedia is a challenging task due to the necessity of data sources that can\nprovide unbiased measurements across multiple countries. Here we show how the\nFacebook Gender Divide (FGD), a metric based on a dataset including more than\n1.4 Billion users in 217 countries, explains various aspects of worldwide\ngender inequality. Our analysis shows that the FGD encodes gender equality\nindices in education, health, and economic opportunity. We find network effects\nthat suggest that using social media has an added value for women. Furthermore,\nwe find that low values of the FGD precede the approach of countries towards\neconomic gender equality. Our results suggest that online social networks,\nwhile suffering evident gender imbalance, may lower the barriers that women\nhave to access informational resources and help to narrow the economic gender\ngap.\n",
"title": "Facebook's gender divide"
}
| null | null | null | null | true | null |
14179
| null |
Default
| null | null |
null |
{
"abstract": " Cooperation is the cornerstone of human evolutionary success. Like no other\nspecies, we champion the sacrifice of personal benefits for the common good,\nand we work together to achieve what we are unable to achieve alone. Knowledge\nand information from past generations is thereby often instrumental in ensuring\nwe keep cooperating rather than deteriorating to less productive ways of\ncoexistence. Here we present a mathematical model based on evolutionary game\ntheory that shows how using the past as the benchmark for evolutionary success,\nrather than just current performance, significantly improves cooperation in the\nfuture. Interestingly, the details of just how the past is taken into account\nplay only second-order importance, whether it be a weighted average of past\npayoffs or just a single payoff value from the past. Cooperation is promoted\nbecause information from the past disables fast invasions of defectors, thus\nenhancing the long-term benefits of cooperative behavior.\n",
"title": "Knowing the past improves cooperation in the future"
}
| null | null | null | null | true | null |
14180
| null |
Default
| null | null |
null |
{
"abstract": " We show that the classical equivalence between the BMO norm and the $L^2$\nnorm of a lacunary Fourier series has an analogue on any discrete group $G$\nequipped with a conditionally negative function.\n",
"title": "BMO estimate of lacunary Fourier series on nonabelian discrete groups"
}
| null | null |
[
"Mathematics"
] | null | true | null |
14181
| null |
Validated
| null | null |
null |
{
"abstract": " Recent literature has demonstrated promising results for training Generative\nAdversarial Networks by employing a set of discriminators, in contrast to the\ntraditional game involving one generator against a single adversary. Such\nmethods perform single-objective optimization on some simple consolidation of\nthe losses, e.g. an arithmetic average. In this work, we revisit the\nmultiple-discriminator setting by framing the simultaneous minimization of\nlosses provided by different models as a multi-objective optimization problem.\nSpecifically, we evaluate the performance of multiple gradient descent and the\nhypervolume maximization algorithm on a number of different datasets. Moreover,\nwe argue that the previously proposed methods and hypervolume maximization can\nall be seen as variations of multiple gradient descent in which the update\ndirection can be computed efficiently. Our results indicate that hypervolume\nmaximization presents a better compromise between sample quality and\ncomputational cost than previous methods.\n",
"title": "Multi-objective training of Generative Adversarial Networks with multiple discriminators"
}
| null | null | null | null | true | null |
14182
| null |
Default
| null | null |
null |
{
"abstract": " We conduct a comprehensive set of tests of performance of surface coils used\nfor nuclear magnetic resonance (NMR) study of quasi 2-dimensional samples. We\nreport ${^{115} \\rm{In}}$ and ${^{31} \\rm{P}}$ NMR measurements on InP,\nsemi-conducting thin substrate samples. Surface coils of both zig-zag\nmeander-line and concentric spiral geometries were used. We compare reception\nsensitivity and signal-to-noise ratio (SNR) of NMR signal obtained by using\nsurface-type coils to that obtained by standard solenoid-type coils. As\nexpected, we find that surface-type coils provide better sensitivity for NMR\nstudy of thin films samples. Moreover, we compare the reception sensitivity of\ndifferent types of the surface coils. We identify the optimal geometry of the\nsurface coils for a given application and/or direction of the applied magnetic\nfield.\n",
"title": "Application of Surface Coil for Nuclear Magnetic Resonance Studies of Semi-conducting Thin Films"
}
| null | null | null | null | true | null |
14183
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we construct an equivariant coarse homology theory with values\nin the category of non-commutative motives of Blumberg, Gepner and Tabuada,\nwith coefficients in any small additive category. Equivariant coarse K-theory\nis obtained from the latter by passing to global sections. The present\nconstruction extends joint work of the first named author with Engel,\nKasprowski and Winges by promoting codomain of the equivariant coarse\nK-homology functor to non-commutative motives.\n",
"title": "A universal coarse K-theory"
}
| null | null | null | null | true | null |
14184
| null |
Default
| null | null |
null |
{
"abstract": " This paper is a continuation of our recent paper devoted to refining the\nparameters of three component (bulge, disk, halo) axisymmetric model Galactic\ngravitational potentials differing by the expression for the dark matter halo\nusing the velocities of distant objects. In all models the bulge and disk\npotentials are described by the Miyamoto-Nagai expressions. In our previous\npaper we used the Allen-Santill'an (I), Wilkinson--Evans (II), and\nNavarro-Frenk-White (III) models to describe the halo. In this paper we use a\nspherical logarithmic Binney potential (model IV), a Plummer sphere (model V),\nand a Hernquist potential (model VI) to describe the halo. A set of present-day\nobservational data in the range of Galactocentric distances R from 0 to 200 kpc\nis used to refine the parameters of the listed models, which are employed most\ncommonly at present. The model rotation curves are fitted to the observed\nvelocities by taking into account the constraints on the local matter density\nand the vertical force . Model VI looks best among the three models considered\nhere from the viewpoint of the achieved accuracy of fitting the model rotation\ncurves to the measurements. This model is close to the Navarro-Frenk-White\nmodel III refined and considered best in our previous paper, which is shown\nusing the integration of the orbits of two globular clusters, Lynga 7 and NGC\n5053, as an example.\n",
"title": "Parameters of Three Selected Model Galactic Potentials Based on the Velocities of Objects at Distances up to 200 kpc"
}
| null | null |
[
"Physics"
] | null | true | null |
14185
| null |
Validated
| null | null |
null |
{
"abstract": " Multilayer MoS2 possesses highly anisotropic thermal conductivities along\nin-plane and cross-plane directions that could hamper heat dissipation in\nelectronics. With about 9% cross-plane compressive strain created by\nhydrostatic pressure in a diamond anvil cell, we observed about 12 times\nincrease in the cross-plane thermal conductivity of multilayer MoS2. Our\nexperimental and theoretical studies reveal that this drastic change arises\nfrom the greatly strengthened interlayer interaction and heavily modified\nphonon dispersions along cross-plane direction, with negligible contribution\nfrom electronic thermal conductivity, despite its enhancement of 4 orders of\nmagnitude. The anisotropic thermal conductivity in the multilayer MoS2 at\nambient environment becomes almost isotropic under highly compressive strain,\neffectively transitioning from 2D to 3D heat dissipation. This strain tuning\napproach also makes possible parallel tuning of structural, thermal and\nelectrical properties, and can be extended to the whole family of 2D Van der\nWaals solids, down to two layer systems.\n",
"title": "Giant Thermal Conductivity Enhancement in Multilayer MoS2 under Highly Compressive Strain"
}
| null | null |
[
"Physics"
] | null | true | null |
14186
| null |
Validated
| null | null |
null |
{
"abstract": " A novel surface interrogation technique is proposed to compute the\nintersection of curves with spline surfaces in isogeometric analysis. The\nintersection points are determined in one-shot without resorting to a\nNewton-Raphson iteration or successive refinement. Surface-curve intersection\nrequires usually the solution of a system of nonlinear equations. It is assumed\nthat the surface is given in form of a spline, such as a NURBS, T-spline or\nCatmull-Clark subdivision surface, and is convertible into a collection of\nBézier patches. First, a hierarchical bounding volume tree is used to\nefficiently identify the Bézier patches with a convex-hull intersecting the\nconvex-hull of a given curve segment. For ease of implementation convex-hulls\nare approximated with k-dops (discrete orientation polytopes). Subsequently,\nthe intersections of the identified Bézier patches with the curve segment are\ndetermined with a matrix-based implicit representation leading to the\ncomputation of a sequence of small singular value decompositions (SVDs). As an\napplication of the developed interrogation technique the isogeometric design\nand analysis of lattice-skin structures is investigated. Current additive\nmanufacturing technologies make it possible to produce up to metre size parts\nwith designed geometric features reaching down to submillimetre scale. The skin\nis a spline surface that is usually created in a computer-aided design (CAD)\nsystem and the periodic lattice to be fitted consists of unit cells, each\ncontaining a small number of struts. The lattice-skin structure is generated by\nprojecting selected lattice nodes onto the surface after determining the\nintersection of unit cell edges with the surface. For mechanical analysis, the\nskin is modelled as a Kirchhoff-Love thin-shell and the lattice as a\npin-jointed truss. The two types of structures are coupled with a standard\nLagrange multiplier approach.\n",
"title": "Interrogation of spline surfaces with application to isogeometric design and analysis of lattice-skin structures"
}
| null | null | null | null | true | null |
14187
| null |
Default
| null | null |
null |
{
"abstract": " We propose general computational procedures based on descriptor state-space\nrealizations to compute coprime factorizations of rational matrices with\nminimum degree denominators. Enhanced recursive pole dislocation techniques are\ndeveloped, which allow to successively place all poles of the factors into a\ngiven \"good\" domain of the complex plane. The resulting McMillan degree of the\ndenominator factor is equal to the number of poles lying in the complementary\n\"bad\" region and therefore is minimal. The new pole dislocation techniques are\nemployed to compute coprime factorizations with proper and stable factors of\narbitrary improper rational matrices and coprime factorizations with inner\ndenominators. The proposed algorithms work for arbitrary descriptor\nrepresentations, regardless they are stabilizable or detectable.\n",
"title": "On recursive computation of coprime factorizations of rational matrices"
}
| null | null | null | null | true | null |
14188
| null |
Default
| null | null |
null |
{
"abstract": " We study electroweak scale Dark Matter (DM) whose interactions with baryonic\nmatter are mediated by a heavy anomalous $Z'$. We emphasize that when the DM is\na Majorana particle, its low-velocity annihilations are dominated by loop\nsuppressed annihilations into the gauge bosons, rather than by p-wave or\nchirally suppressed annihilations into the SM fermions. Because the $Z'$ is\nanomalous, these kinds of DM models can be realized only as effective field\ntheories (EFTs) with a well-defined cutoff, where heavy spectator fermions\nrestore gauge invariance at high energies. We formulate these EFTs, estimate\ntheir cutoff and properly take into account the effect of the Chern-Simons\nterms one obtains after the spectator fermions are integrated out. We find\nthat, while for light DM collider and direct detection experiments usually\nprovide the strongest bounds, the bounds at higher masses are heavily dominated\nby indirect detection experiments, due to strong annihilation into $W^+W^-$,\n$ZZ$, $Z\\gamma$ and possibly into $gg$ and $\\gamma\\gamma$. We emphasize that\nthese annihilation channels are generically significant because of the\nstructure of the EFT, and therefore these models are prone to strong indirect\ndetection constraints. Even though we focus on selected $Z'$ models for\nillustrative purposes, our setup is completely generic and can be used for\nanalyzing the predictions of any anomalous $Z'$-mediated DM model with\narbitrary charges.\n",
"title": "On Dark Matter Interactions with the Standard Model through an Anomalous $Z'$"
}
| null | null | null | null | true | null |
14189
| null |
Default
| null | null |
null |
{
"abstract": " Generalised hydrodynamics predicts universal ballistic transport in\nintegrable lattice systems when prepared in generic inhomogeneous initial\nstates. However, the ballistic contribution to transport can vanish in systems\nwith additional discrete symmetries. Here we perform large scale numerical\nsimulations of spin dynamics in the anisotropic Heisenberg $XXZ$ spin $1/2$\nchain starting from an inhomogeneous mixed initial state which is symmetric\nwith respect to a combination of spin-reversal and spatial reflection. In the\nisotropic and easy-axis regimes we find non-ballistic spin transport which we\nanalyse in detail in terms of scaling exponents of the transported\nmagnetisation and scaling profiles of the spin density. While in the easy-axis\nregime we find accurate evidence of normal diffusion, the spin transport in the\nisotropic case is clearly super-diffusive, with the scaling exponent very close\nto $2/3$, but with universal scaling dynamics which obeys the diffusion\nequation in nonlinearly scaled time.\n",
"title": "Spin diffusion from an inhomogeneous quench in an integrable system"
}
| null | null | null | null | true | null |
14190
| null |
Default
| null | null |
null |
{
"abstract": " We propose to employ scale spaces of mathematical morphology to\nhierarchically simplify fracture surfaces of complementarily fitting\narchaeological fragments. This representation preserves contact and is\ninsensitive to different kinds of abrasion affecting the exact complementarity\nof the original fragments. We present a pipeline for morphologically\nsimplifying fracture surfaces, based on their Lipschitz nature; its core is a\nnew embedding of fracture surfaces to simultaneously compute both closing and\nopening morphological operations, using distance transforms.\n",
"title": "Morphological Simplification of Archaeological Fracture Surfaces"
}
| null | null | null | null | true | null |
14191
| null |
Default
| null | null |
null |
{
"abstract": " The paper deals with a construction of a separating system of rational\ninvariants for finite dimensional generic algebras. In the process of dealing\nan approach to a rough classification of finite dimensional algebras is offered\nby attaching them some quadratic forms.\n",
"title": "A note on a separating system of rational invariants for finite dimensional generic algebras"
}
| null | null | null | null | true | null |
14192
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we present a method for simultaneously segmenting brain tumors\nand an extensive set of organs-at-risk for radiation therapy planning of\nglioblastomas. The method combines a contrast-adaptive generative model for\nwhole-brain segmentation with a new spatial regularization model of tumor shape\nusing convolutional restricted Boltzmann machines. We demonstrate\nexperimentally that the method is able to adapt to image acquisitions that\ndiffer substantially from any available training data, ensuring its\napplicability across treatment sites; that its tumor segmentation accuracy is\ncomparable to that of the current state of the art; and that it captures most\norgans-at-risk sufficiently well for radiation therapy planning purposes. The\nproposed method may be a valuable step towards automating the delineation of\nbrain tumors and organs-at-risk in glioblastoma patients undergoing radiation\ntherapy.\n",
"title": "A Modality-Adaptive Method for Segmenting Brain Tumors and Organs-at-Risk in Radiation Therapy Planning"
}
| null | null | null | null | true | null |
14193
| null |
Default
| null | null |
null |
{
"abstract": " Protein pattern formation is essential for the spatial organization of many\nintracellular processes like cell division, flagellum positioning, and\nchemotaxis. A prominent example of intracellular patterns are the oscillatory\npole-to-pole oscillations of Min proteins in \\textit{E. coli} whose biological\nfunction is to ensure precise cell division. Cell polarization, a prerequisite\nfor processes such as stem cell differentiation and cell polarity in yeast, is\nalso mediated by a diffusion-reaction process. More generally, these functional\nmodules of cells serve as model systems for self-organization, one of the core\nprinciples of life. Under which conditions spatio-temporal patterns emerge, and\nhow these patterns are regulated by biochemical and geometrical factors are\nmajor aspects of current research. Here we review recent theoretical and\nexperimental advances in the field of intracellular pattern formation, focusing\non general design principles and fundamental physical mechanisms.\n",
"title": "Protein Pattern Formation"
}
| null | null | null | null | true | null |
14194
| null |
Default
| null | null |
null |
{
"abstract": " Artificial intelligence (AI) generally and machine learning (ML) specifically\ndemonstrate impressive practical success in many different application domains,\ne.g. in autonomous driving, speech recognition, or recommender systems. Deep\nlearning approaches, trained on extremely large data sets or using\nreinforcement learning methods have even exceeded human performance in visual\ntasks, particularly on playing games such as Atari, or mastering the game of\nGo. Even in the medical domain there are remarkable results. The central\nproblem of such models is that they are regarded as black-box models and even\nif we understand the underlying mathematical principles, they lack an explicit\ndeclarative knowledge representation, hence have difficulty in generating the\nunderlying explanatory structures. This calls for systems enabling to make\ndecisions transparent, understandable and explainable. A huge motivation for\nour approach are rising legal and privacy aspects. The new European General\nData Protection Regulation entering into force on May 25th 2018, will make\nblack-box approaches difficult to use in business. This does not imply a ban on\nautomatic learning approaches or an obligation to explain everything all the\ntime, however, there must be a possibility to make the results re-traceable on\ndemand. In this paper we outline some of our research topics in the context of\nthe relatively new area of explainable-AI with a focus on the application in\nmedicine, which is a very special domain. This is due to the fact that medical\nprofessionals are working mostly with distributed heterogeneous and complex\nsources of data. In this paper we concentrate on three sources: images, *omics\ndata and text. We argue that research in explainable-AI would generally help to\nfacilitate the implementation of AI/ML in the medical domain, and specifically\nhelp to facilitate transparency and trust.\n",
"title": "What do we need to build explainable AI systems for the medical domain?"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14195
| null |
Validated
| null | null |
null |
{
"abstract": " Matrix completion is a problem that arises in many data-analysis settings\nwhere the input consists of a partially-observed matrix (e.g., recommender\nsystems, traffic matrix analysis etc.). Classical approaches to matrix\ncompletion assume that the input partially-observed matrix is low rank. The\nsuccess of these methods depends on the number of observed entries and the rank\nof the matrix; the larger the rank, the more entries need to be observed in\norder to accurately complete the matrix. In this paper, we deal with matrices\nthat are not necessarily low rank themselves, but rather they contain low-rank\nsubmatrices. We propose Targeted, which is a general framework for completing\nsuch matrices. In this framework, we first extract the low-rank submatrices and\nthen apply a matrix-completion algorithm to these low-rank submatrices as well\nas the remainder matrix separately. Although for the completion itself we use\nstate-of-the-art completion methods, our results demonstrate that Targeted\nachieves significantly smaller reconstruction errors than other classical\nmatrix-completion methods. One of the key technical contributions of the paper\nlies in the identification of the low-rank submatrices from the input\npartially-observed matrices.\n",
"title": "Targeted matrix completion"
}
| null | null | null | null | true | null |
14196
| null |
Default
| null | null |
null |
{
"abstract": " Many applications infer the structure of a probabilistic graphical model from\ndata to elucidate the relationships between variables. But how can we train\ngraphical models on a massive data set? In this paper, we show how to construct\ncoresets -compressed data sets which can be used as proxy for the original data\nand have provably bounded worst case error- for Gaussian dependency networks\n(DNs), i.e., cyclic directed graphical models over Gaussians, where the parents\nof each variable are its Markov blanket. Specifically, we prove that Gaussian\nDNs admit coresets of size independent of the size of the data set.\nUnfortunately, this does not extend to DNs over members of the exponential\nfamily in general. As we will prove, Poisson DNs do not admit small coresets.\nDespite this worst-case result, we will provide an argument why our coreset\nconstruction for DNs can still work well in practice on count data. To\ncorroborate our theoretical results, we empirically evaluated the resulting\nCore DNs on real data sets. The results\n",
"title": "Coresets for Dependency Networks"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
14197
| null |
Validated
| null | null |
null |
{
"abstract": " The influence of the B-site ion substitutions in\n(1-x)(Bi1/2Na1/2)TiO3-xBaTiO3 system of solid solutions on the relative\nstability of the ferroelectric and antiferroelectric phases has been studied.\nThe ions of zirconium, tin, along with (In0.5Nb0.5), (Fe0.5Nb0.5), (Al0.5V0.5)\nion complexes have been used as substituting elements. An increase in the\nconcentration of the substituting ion results in a near linear variation in the\nsize of the crystal lattice cell. Along with the cell size variation a change\nin the relative stability of the ferroelectric and antiferroelectric phases\ntakes place according to the changes of the tolerance factor of the solid\nsolution. An increase in the tolerance factor leads to the increase in the\ntemperature of the ferroelectric-antiferroelectric phase transition, and vice\nversa. All obtained results demonstrate the predominant influence of the ion\nsize factor on the relative stability of the ferroelectric and\nantiferroelectric states in the (Na0.5Bi0.5)TiO3-based solid solutions and\nindicate the way for raising the temperature of the\nferroelectric-antiferroelectric phase transition.\n",
"title": "Relative stability of a ferroelectric state in (Na0.5Bi0.5)TiO3-based compounds under substitutions: Role of a tolerance factor in expansion of the temperature interval of stable ferroelectric state"
}
| null | null |
[
"Physics"
] | null | true | null |
14198
| null |
Validated
| null | null |
null |
{
"abstract": " This paper presents an exhaustive study on the arrivals process at eight\nimportant European airports. Using inbound traffic data, we define, compare,\nand contrast a data-driven Poisson and PSRA point process. Although, there is\nsufficient evidence that the interarrivals might follow an exponential\ndistribution, this finding does not directly translate to evidence that the\narrivals stream is Poisson. The main reason is that finite-capacity constraints\nimpose a correlation structure to the arrivals stream, which a Poisson model\ncannot capture. We show the weaknesses and somehow the difficulties of using a\nPoisson process to model with good approximation the arrivals stream. On the\nother hand, our innovative non-parametric, data-driven PSRA model, predicts\nquite well and captures important properties of the typical arrivals stream.\n",
"title": "Data-driven modelling and validation of aircraft inbound-stream at some major European airports"
}
| null | null | null | null | true | null |
14199
| null |
Default
| null | null |
null |
{
"abstract": " We derive a new exact evolution equation for the scale dependence of an\neffective action. The corresponding equation for the effective potential\npermits a useful truncation. This allows one to deal with the infrared problems\nof theories with massless modes in less than four dimensions which are relevant\nfor the high temperature phase transition in particle physics or the\ncomputation of critical exponents in statistical mechanics.\n",
"title": "Exact evolution equation for the effective potential"
}
| null | null |
[
"Physics"
] | null | true | null |
14200
| null |
Validated
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.