text
null | inputs
dict | prediction
null | prediction_agent
null | annotation
list | annotation_agent
null | multi_label
bool 1
class | explanation
null | id
stringlengths 1
5
| metadata
null | status
stringclasses 2
values | event_timestamp
null | metrics
null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
null |
{
"abstract": " We extend existing methods for using cross-correlations to derive redshift\ndistributions for photometric galaxies, without using photometric redshifts.\nThe model presented in this paper simultaneously yields highly accurate and\nunbiased redshift distributions and, for the first time, redshift-dependent\nluminosity functions, using only clustering information and the apparent\nmagnitudes of the galaxies as input. In contrast to many existing techniques\nfor recovering unbiased redshift distributions, the output of our method is not\ndegenerate with the galaxy bias b(z), which is achieved by modelling the shape\nof the luminosity bias. We successfully apply our method to a mock galaxy\nsurvey and discuss improvements to be made before applying our model to real\ndata.\n",
"title": "A cross-correlation-based estimate of the galaxy luminosity function"
}
| null | null | null | null | true | null |
13301
| null |
Default
| null | null |
null |
{
"abstract": " We propose a novel block-row partitioning method in order to improve the\nconvergence rate of the block Cimmino algorithm for solving general sparse\nlinear systems of equations. The convergence rate of the block Cimmino\nalgorithm depends on the orthogonality among the block rows obtained by the\npartitioning method. The proposed method takes numerical orthogonality among\nblock rows into account by proposing a row inner-product graph model of the\ncoefficient matrix. In the graph partitioning formulation defined on this graph\nmodel, the partitioning objective of minimizing the cutsize directly\ncorresponds to minimizing the sum of inter-block inner products between block\nrows thus leading to an improvement in the eigenvalue spectrum of the iteration\nmatrix. This in turn leads to a significant reduction in the number of\niterations required for convergence. Extensive experiments conducted on a large\nset of matrices confirm the validity of the proposed method against a\nstate-of-the-art method.\n",
"title": "A Novel Partitioning Method for Accelerating the Block Cimmino Algorithm"
}
| null | null | null | null | true | null |
13302
| null |
Default
| null | null |
null |
{
"abstract": " Many image processing tasks involve image-to-image mapping, which can be\naddressed well by fully convolutional networks (FCN) without any heavy\npreprocessing. Although empirically designing and training FCNs can achieve\nsatisfactory results, reasons for the improvement in performance are slightly\nambiguous. Our study is to make progress in understanding their generalization\nabilities through visualizing the optimization landscapes. The visualization of\nobjective functions is obtained by choosing a solution and projecting its\nvicinity onto a 3D space. We compare three FCN-based networks (two existing\nmodels and a new proposed in this paper for comparison) on multiple datasets.\nIt has been observed in practice that the connections from the pre-pooled\nfeature maps to the post-upsampled can achieve better results. We investigate\nthe cause and provide experiments to shows that the skip-layer connections in\nFCN can promote flat optimization landscape, which is well known to generalize\nbetter. Additionally, we explore the relationship between the models\ngeneralization ability and loss surface under different batch sizes. Results\nshow that large-batch training makes the model converge to sharp minimizers\nwith chaotic vicinities while small-batch method leads the model to flat\nminimizers with smooth and nearly convex regions. Our work may contribute to\ninsights and analysis for designing and training FCNs.\n",
"title": "Visualized Insights into the Optimization Landscape of Fully Convolutional Networks"
}
| null | null | null | null | true | null |
13303
| null |
Default
| null | null |
null |
{
"abstract": " Exact lower and upper bounds on the best possible misclassification\nprobability for a finite number of classes are obtained in terms of the total\nvariation norms of the differences between the sub-distributions over the\nclasses. These bounds are compared with the exact bounds in terms of the\nconditional entropy obtained by Feder and Merhav.\n",
"title": "Exact upper and lower bounds on the misclassification probability"
}
| null | null | null | null | true | null |
13304
| null |
Default
| null | null |
null |
{
"abstract": " The big graph database model provides strong modeling for complex\napplications and efficient querying. However, it is still a big challenge to\nfind all exact matches of a query graph in a big graph database, which is known\nas the subgraph isomorphism problem. The current subgraph isomorphism\napproaches are built on Ullmann's idea of focusing on the strategy of pruning\nout the irrelevant candidates. Nevertheless, the existing pruning techniques\nneed much more improvement to efficiently handle complex queries. Moreover,\nmany of those existing algorithms need large indices requiring extra memory\nconsumption. Motivated by these, we introduce a new subgraph isomorphism\nalgorithm, named as BB-Graph, for querying big graph databases efficiently\nwithout requiring a large data structure to be stored in main memory. We test\nand compare our proposed BB-Graph algorithm with two popular existing\napproaches, GraphQL and Cypher. Our experiments are done on three different\ndata sets; (1) a very big graph database of a real-life population database,\n(2) a graph database of a simulated bank database, and (3) the publicly\navailable World Cup big graph database. We show that our solution performs\nbetter than those algorithms mentioned here for most of the query types\nexperimented on these big databases.\n",
"title": "BB-Graph: A Subgraph Isomorphism Algorithm for Efficiently Querying Big Graph Databases"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13305
| null |
Validated
| null | null |
null |
{
"abstract": " We present analytic self-similar solutions for the one, two and three\ndimensional Madelung hydrodynamical equation for a free particle. There is a\ndirect connection between the zeros of the Madelung fluid density and the\nmagnitude of the quantum potential.\n",
"title": "Analytic solutions of the Madelung equation"
}
| null | null | null | null | true | null |
13306
| null |
Default
| null | null |
null |
{
"abstract": " We employ unsupervised machine learning techniques to learn latent parameters\nwhich best describe states of the two-dimensional Ising model and the\nthree-dimensional XY model. These methods range from principal component\nanalysis to artificial neural network based variational autoencoders. The\nstates are sampled using a Monte-Carlo simulation above and below the critical\ntemperature. We find that the predicted latent parameters correspond to the\nknown order parameters. The latent representation of the states of the models\nin question are clustered, which makes it possible to identify phases without\nprior knowledge of their existence or the underlying Hamiltonian. Furthermore,\nwe find that the reconstruction loss function can be used as a universal\nidentifier for phase transitions.\n",
"title": "Unsupervised learning of phase transitions: from principal component analysis to variational autoencoders"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
13307
| null |
Validated
| null | null |
null |
{
"abstract": " Design of energy-efficient access networks has emerged as an important area\nof research, since access networks consume $80-90\\%$ of the overall Internet\npower consumption. TWDM-PON is envisaged to be one of the widely accepted\nfuture access technologies. TWDM-PON offers an additional opportunity to save\nenergy at the OLT along with the existing energy-efficient ONU design. In this\npaper, we focus on the energy-efficient OLT design in a TWDM-PON. While most of\nthe conventional methods employ a minimization of the number of wavelengths, we\npropose a novel approach which aims at minimizing the number of voids created\ndue to scheduling. In the process, for the first time, we present a\nlow-complexity on-line scheduling algorithm for the upstream traffic\nconsidering delay constraints. Our extensive simulations demonstrate a\nsignificant improvement in energy efficiency of $\\sim 25\\%$ for high load at\nthe OLT receivers. Furthermore, we provide an analytical upper-bound on the\nenergy-efficiency of the OLT receivers and demonstrate that the proposed\nprotocol achieves an energy efficiency very close to the bound with a maximum\ndeviation $\\sim 2\\%$ for $64$ ONUs.\n",
"title": "A novel online scheduling protocol for energy-efficient TWDM-OLT design"
}
| null | null | null | null | true | null |
13308
| null |
Default
| null | null |
null |
{
"abstract": " I discuss the evolution of computer architectures with a focus on QCD and\nwith reference to the interplay between architecture, engineering, data motion\nand algorithms. New architectures are discussed and recent performance results\nare displayed. I also review recent progress in multilevel solver and\nintegation algorithms.\n",
"title": "Machines and Algorithms"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
13309
| null |
Validated
| null | null |
null |
{
"abstract": " Person re-identification (Re-ID) aims at matching images of the same person\nacross disjoint camera views, which is a challenging problem in multimedia\nanalysis, multimedia editing and content-based media retrieval communities. The\nmajor challenge lies in how to preserve similarity of the same person across\nvideo footages with large appearance variations, while discriminating different\nindividuals. To address this problem, conventional methods usually consider the\npairwise similarity between persons by only measuring the point to point (P2P)\ndistance. In this paper, we propose to use deep learning technique to model a\nnovel set to set (S2S) distance, in which the underline objective focuses on\npreserving the compactness of intra-class samples for each camera view, while\nmaximizing the margin between the intra-class set and inter-class set. The S2S\ndistance metric is consisted of three terms, namely the class-identity term,\nthe relative distance term and the regularization term. The class-identity term\nkeeps the intra-class samples within each camera view gathering together, the\nrelative distance term maximizes the distance between the intra-class class set\nand inter-class set across different camera views, and the regularization term\nsmoothness the parameters of deep convolutional neural network (CNN). As a\nresult, the final learned deep model can effectively find out the matched\ntarget to the probe object among various candidates in the video gallery by\nlearning discriminative and stable feature representations. Using the CUHK01,\nCUHK03, PRID2011 and Market1501 benchmark datasets, we extensively conducted\ncomparative evaluations to demonstrate the advantages of our method over the\nstate-of-the-art approaches.\n",
"title": "Large Margin Learning in Set to Set Similarity Comparison for Person Re-identification"
}
| null | null | null | null | true | null |
13310
| null |
Default
| null | null |
null |
{
"abstract": " Bounded model checking is among the most efficient techniques for the\nautomatic verification of concurrent programs. However, encoding all possible\ninterleavings often requires a huge and complex formula, which significantly\nlimits the salability. This paper proposes a novel and efficient abstraction\nrefinement method for multi-threaded program verification. Observing that the\nhuge formula is usually dominated by the exact encoding of the scheduling\nconstraint, this paper proposes a \\tsc based abstraction refinement method,\nwhich avoids the huge and complex encoding of BMC. In addition, to obtain an\neffective refinement, we have devised two graph-based algorithms over event\norder graph for counterexample validation and refinement generation, which can\nalways obtain a small yet effective refinement constraint. Enhanced by two\nconstraint-based algorithms for counterexample validation and refinement\ngeneration, we have proved that our method is sound and complete w.r.t. the\ngiven loop unwinding depth. Experimental results on \\svcompc benchmarks\nindicate that our method is promising and significantly outperforms the\nexisting state-of-the-art tools.\n",
"title": "Scheduling Constraint Based Abstraction Refinement for Multi-Threaded Program Verification"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13311
| null |
Validated
| null | null |
null |
{
"abstract": " We survey and compare various generalizations of braid groups for quivers\nwith superpotential and focus on the cluster braid groups, which are introduced\nin a joint work with A.~King. Our motivations come from the study of cluster\nalgebras, Calabi-Yau categories and Bridgeland stability conditions.\n",
"title": "The braid group for a quiver with superpotential"
}
| null | null | null | null | true | null |
13312
| null |
Default
| null | null |
null |
{
"abstract": " Let $Z=G/H$ be the homogeneous space of a real reductive group and a\nunimodular real spherical subgroup, and consider the regular representation of\n$G$ on $L^2(Z)$. It is shown that all representations of the discrete series,\nthat is, the irreducible subrepresentations of $L^2(Z)$, have infinitesimal\ncharacters which are real and belong to a lattice. Moreover, let $K$ be a\nmaximal compact subgroup of $G$. Then each irreducible representation of $K$\noccurs in a finite set of such discrete series representations only. Similar\nresults are obtained for the twisted discrete series, that is, the discrete\ncomponents of the space of square integrable sections of a line bundle, given\nby a unitary character on an abelian extension of $H$.\n",
"title": "The infinitesimal characters of discrete series for real spherical spaces"
}
| null | null | null | null | true | null |
13313
| null |
Default
| null | null |
null |
{
"abstract": " The purpose of this work is to introduce a general class of $C_G$-simulation\nfunctions and obtained some new coincidence and common fixed points results in\nmetric spaces. Some useful examples are presented to illustrate our theorems.\nResults obtained in this paper extend, generalize and unify some well known\nfixed and common fixed point results.\n",
"title": "Coincidence point results involving a generalized class of simulation functions"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13314
| null |
Validated
| null | null |
null |
{
"abstract": " State-sponsored \"bad actors\" increasingly weaponize social media platforms to\nlaunch cyberattacks and disinformation campaigns during elections. Social media\ncompanies, due to their rapid growth and scale, struggle to prevent the\nweaponization of their platforms. This study conducts an automated spear\nphishing and disinformation campaign on Twitter ahead of the 2018 United States\nMidterm Elections. A fake news bot account - the @DCNewsReport - was created\nand programmed to automatically send customized tweets with a \"breaking news\"\nlink to 138 Twitter users, before being restricted by Twitter.\nOverall, one in five users clicked the link, which could have potentially led\nto the downloading of ransomware or the theft of private information. However,\nthe link in this experiment was non-malicious and redirected users to a Google\nForms survey. In predicting users' likelihood to click the link on Twitter, no\nstatistically significant differences were observed between right-wing and\nleft-wing partisans, or between Web users and mobile users. The findings signal\nthat politically expressive Americans on Twitter, regardless of their party\npreferences or the devices they use to access the platform, are at risk of\nbeing spear phishing on social media.\n",
"title": "A Simulated Cyberattack on Twitter: Assessing Partisan Vulnerability to Spear Phishing and Disinformation ahead of the 2018 U.S. Midterm Elections"
}
| null | null | null | null | true | null |
13315
| null |
Default
| null | null |
null |
{
"abstract": " Estimates for asteroid masses are based on their gravitational perturbations\non the orbits of other objects such as Mars, spacecraft, or other asteroids\nand/or their satellites. In the case of asteroid-asteroid perturbations, this\nleads to an inverse problem in at least 13 dimensions where the aim is to\nderive the mass of the perturbing asteroid(s) and six orbital elements for both\nthe perturbing asteroid(s) and the test asteroid(s) based on astrometric\nobservations. We have developed and implemented three different mass estimation\nalgorithms utilizing asteroid-asteroid perturbations: the very rough 'marching'\napproximation, in which the asteroids' orbital elements are not fitted, thereby\nreducing the problem to a one-dimensional estimation of the mass, an\nimplementation of the Nelder-Mead simplex method, and most significantly, a\nMarkov-chain Monte Carlo (MCMC) approach. We describe each of these algorithms\nwith particular focus on the MCMC algorithm, and present example results using\nboth synthetic and real data. Our results agree with the published mass\nestimates, but suggest that the published uncertainties may be misleading as a\nconsequence of using linearized mass-estimation methods. Finally, we discuss\nremaining challenges with the algorithms as well as future plans.\n",
"title": "Asteroid mass estimation using Markov-chain Monte Carlo"
}
| null | null |
[
"Physics"
] | null | true | null |
13316
| null |
Validated
| null | null |
null |
{
"abstract": " Markov regime switching models have been widely used in numerous empirical\napplications in economics and finance. However, the asymptotic distribution of\nthe maximum likelihood estimator (MLE) has not been proven for some empirically\npopular Markov regime switching models. In particular, the asymptotic\ndistribution of the MLE has been unknown for models in which some elements of\nthe transition probability matrix have the value of zero, as is commonly\nassumed in empirical applications with models with more than two regimes. This\nalso includes models in which the regime-specific density depends on both the\ncurrent and the lagged regimes such as the seminal model of Hamilton (1989) and\nswitching ARCH model of Hamilton and Susmel (1994). This paper shows the\nasymptotic normality of the MLE and consistency of the asymptotic covariance\nmatrix estimate of these models.\n",
"title": "Asymptotic Properties of the Maximum Likelihood Estimator in Regime Switching Econometric Models"
}
| null | null | null | null | true | null |
13317
| null |
Default
| null | null |
null |
{
"abstract": " The VIPAFLEET project consists in developing models and algorithms for man-\naging a fleet of Individual Public Autonomous Vehicles (VIPA). Hereby, we\nconsider a fleet of cars distributed at specified stations in an industrial\narea to supply internal transportation, where the cars can be used in different\nmodes of circulation (tram mode, elevator mode, taxi mode). One goal is to\ndevelop and implement suitable algorithms for each mode in order to satisfy all\nthe requests under an economic point of view by minimizing the total tour\nlength. The innovative idea and challenge of the project is to develop and\ninstall a dynamic fleet management system that allows the operator to switch\nbetween the different modes within the different periods of the day according\nto the dynamic transportation demands of the users. We model the underlying\nonline transportation system and propose a correspond- ing fleet management\nframework, to handle modes, demands and commands. We consider two modes of\ncirculation, tram and elevator mode, propose for each mode appropriate on- line\nalgorithms and evaluate their performance, both in terms of competitive\nanalysis and practical behavior.\n",
"title": "Fleet management for autonomous vehicles: Online PDP under special constraints"
}
| null | null | null | null | true | null |
13318
| null |
Default
| null | null |
null |
{
"abstract": " While the Bayesian Information Criterion (BIC) and Akaike Information\nCriterion (AIC) are powerful tools for model selection in linear regression,\nthey are built on different prior assumptions and thereby apply to different\ndata generation scenarios. We show that in the finite-dimensional case their\nrespective assumptions can be unified within an augmented model-plus-noise\nspace and construct a prior in this space which inherits the beneficial\nproperties of both AIC and BIC. This allows us to adapt the BIC to be robust\nagainst misspecified models where the signal to noise ratio is low.\n",
"title": "Bayesian Model Selection for Misspecified Models in Linear Regression"
}
| null | null | null | null | true | null |
13319
| null |
Default
| null | null |
null |
{
"abstract": " We analyze low rank tensor completion (TC) using noisy measurements of a\nsubset of the tensor. Assuming a rank-$r$, order-$d$, $N \\times N \\times \\cdots\n\\times N$ tensor where $r=O(1)$, the best sampling complexity that was achieved\nis $O(N^{\\frac{d}{2}})$, which is obtained by solving a tensor nuclear-norm\nminimization problem. However, this bound is significantly larger than the\nnumber of free variables in a low rank tensor which is $O(dN)$. In this paper,\nwe show that by using an atomic-norm whose atoms are rank-$1$ sign tensors, one\ncan obtain a sample complexity of $O(dN)$. Moreover, we generalize the matrix\nmax-norm definition to tensors, which results in a max-quasi-norm (max-qnorm)\nwhose unit ball has small Rademacher complexity. We prove that solving a\nconstrained least squares estimation using either the convex atomic-norm or the\nnonconvex max-qnorm results in optimal sample complexity for the problem of\nlow-rank tensor completion. Furthermore, we show that these bounds are nearly\nminimax rate-optimal. We also provide promising numerical results for max-qnorm\nconstrained tensor completion, showing improved recovery results compared to\nmatricization and alternating least squares.\n",
"title": "Near-optimal sample complexity for convex tensor completion"
}
| null | null | null | null | true | null |
13320
| null |
Default
| null | null |
null |
{
"abstract": " In recent years much effort has been concentrated towards achieving\npolynomial time lower bounds on algorithms for solving various well-known\nproblems. A useful technique for showing such lower bounds is to prove them\nconditionally based on well-studied hardness assumptions such as 3SUM, APSP,\nSETH, etc. This line of research helps to obtain a better understanding of the\ncomplexity inside P.\nA related question asks to prove conditional space lower bounds on data\nstructures that are constructed to solve certain algorithmic tasks after an\ninitial preprocessing stage. This question received little attention in\nprevious research even though it has potential strong impact.\nIn this paper we address this question and show that surprisingly many of the\nwell-studied hard problems that are known to have conditional polynomial time\nlower bounds are also hard when concerning space. This hardness is shown as a\ntradeoff between the space consumed by the data structure and the time needed\nto answer queries. The tradeoff may be either smooth or admit one or more\nsingularity points.\nWe reveal interesting connections between different space hardness\nconjectures and present matching upper bounds. We also apply these hardness\nconjectures to both static and dynamic problems and prove their conditional\nspace hardness.\nWe believe that this novel framework of polynomial space conjectures can play\nan important role in expressing polynomial space lower bounds of many important\nalgorithmic problems. Moreover, it seems that it can also help in achieving a\nbetter understanding of the hardness of their corresponding problems in terms\nof time.\n",
"title": "Conditional Lower Bounds for Space/Time Tradeoffs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13321
| null |
Validated
| null | null |
null |
{
"abstract": " Cardiac left ventricle (LV) quantification is among the most clinically\nimportant tasks for identification and diagnosis of cardiac diseases, yet still\na challenge due to the high variability of cardiac structure and the complexity\nof temporal dynamics. Full quantification, i.e., to simultaneously quantify all\nLV indices including two areas (cavity and myocardium), six regional wall\nthicknesses (RWT), three LV dimensions, and one cardiac phase, is even more\nchallenging since the uncertain relatedness intra and inter each type of\nindices may hinder the learning procedure from better convergence and\ngeneralization. In this paper, we propose a newly-designed multitask learning\nnetwork (FullLVNet), which is constituted by a deep convolution neural network\n(CNN) for expressive feature embedding of cardiac structure; two followed\nparallel recurrent neural network (RNN) modules for temporal dynamic modeling;\nand four linear models for the final estimation. During the final estimation,\nboth intra- and inter-task relatedness are modeled to enforce improvement of\ngeneralization: 1) respecting intra-task relatedness, group lasso is applied to\neach of the regression tasks for sparse and common feature selection and\nconsistent prediction; 2) respecting inter-task relatedness, three phase-guided\nconstraints are proposed to penalize violation of the temporal behavior of the\nobtained LV indices. Experiments on MR sequences of 145 subjects show that\nFullLVNet achieves high accurate prediction with our intra- and inter-task\nrelatedness, leading to MAE of 190mm$^2$, 1.41mm, 2.68mm for average areas,\nRWT, dimensions and error rate of 10.4\\% for the phase classification. This\nendows our method a great potential in comprehensive clinical assessment of\nglobal, regional and dynamic cardiac function.\n",
"title": "Full Quantification of Left Ventricle via Deep Multitask Learning Network Respecting Intra- and Inter-Task Relatedness"
}
| null | null | null | null | true | null |
13322
| null |
Default
| null | null |
null |
{
"abstract": " This paper investigates properties of the class of graphs based on\nexchangeable point processes. We provide asymptotic expressions for the number\nof edges, number of nodes and degree distributions, identifying four regimes: a\ndense regime, a sparse, almost dense regime, a sparse regime with power-law\nbehavior, and an almost extremely sparse regime. Our results allow us to derive\na consistent estimator for the scalar parameter tuning the sparsity of the\ngraph. We also propose a class of models within this framework where one can\nseparately control the local, latent structure and the global\nsparsity/power-law properties of the graph.\n",
"title": "On sparsity and power-law properties of graphs based on exchangeable point processes"
}
| null | null | null | null | true | null |
13323
| null |
Default
| null | null |
null |
{
"abstract": " Many real-world communication networks often have hybrid nature with both\nfixed nodes and moving modes, such as the mobile phone networks mainly composed\nof fixed base stations and mobile phones. In this paper, we discuss the\ninformation transmission process on the hybrid networks with both fixed and\nmobile nodes. The fixed nodes (base stations) are connected as a spatial\nlattice on the plane forming the information-carrying backbone, while the\nmobile nodes (users), which are the sources and destinations of information\npackets, connect to their current nearest fixed nodes respectively to deliver\nand receive information packets. We observe the phase transition of traffic\nload in the hybrid network when the packet generation rate goes from below and\nthen above a critical value, which measures the network capacity of packets\ndelivery. We obtain the optimal speed of moving nodes leading to the maximum\nnetwork capacity. We further improve the network capacity by rewiring the fixed\nnodes and by considering the current load of fixed nodes during packets\ntransmission. Our purpose is to optimize the network capacity of hybrid\nnetworks from the perspective of network science, and provide some insights for\nthe construction of future communication infrastructures.\n",
"title": "Information transmission on hybrid networks"
}
| null | null | null | null | true | null |
13324
| null |
Default
| null | null |
null |
{
"abstract": " We propose a novel technique to make neural network robust to adversarial\nexamples using a generative adversarial network. We alternately train both\nclassifier and generator networks. The generator network generates an\nadversarial perturbation that can easily fool the classifier network by using a\ngradient of each image. Simultaneously, the classifier network is trained to\nclassify correctly both original and adversarial images generated by the\ngenerator. These procedures help the classifier network to become more robust\nto adversarial perturbations. Furthermore, our adversarial training framework\nefficiently reduces overfitting and outperforms other regularization methods\nsuch as Dropout. We applied our method to supervised learning for CIFAR\ndatasets, and experimantal results show that our method significantly lowers\nthe generalization error of the network. To the best of our knowledge, this is\nthe first method which uses GAN to improve supervised learning.\n",
"title": "Generative Adversarial Trainer: Defense to Adversarial Perturbations with GAN"
}
| null | null | null | null | true | null |
13325
| null |
Default
| null | null |
null |
{
"abstract": " Computations over the rational numbers often suffer from intermediate\ncoefficient swell. One solution to this problem is to apply the given algorithm\nmodulo a number of primes and then lift the modular results to the rationals.\nThis method is guaranteed to work if we use a sufficiently large set of good\nprimes. In many applications, however, there is no efficient way of excluding\nbad primes. In this note, we describe a technique for rational reconstruction\nwhich will nevertheless return the correct result, provided the number of good\nprimes in the selected set of primes is large enough. We give a number of\nillustrating examples which are implemented using the computer algebra system\nSingular and the programming language Julia. We discuss applications of our\ntechnique in computational algebraic geometry.\n",
"title": "Bad Primes in Computational Algebraic Geometry"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
13326
| null |
Validated
| null | null |
null |
{
"abstract": " Collaborative filtering often suffers from sparsity and cold start problems\nin real recommendation scenarios, therefore, researchers and engineers usually\nuse side information to address the issues and improve the performance of\nrecommender systems. In this paper, we consider knowledge graphs as the source\nof side information. We propose MKR, a Multi-task feature learning approach for\nKnowledge graph enhanced Recommendation. MKR is a deep end-to-end framework\nthat utilizes knowledge graph embedding task to assist recommendation task. The\ntwo tasks are associated by cross&compress units, which automatically share\nlatent features and learn high-order interactions between items in recommender\nsystems and entities in the knowledge graph. We prove that cross&compress units\nhave sufficient capability of polynomial approximation, and show that MKR is a\ngeneralized framework over several representative methods of recommender\nsystems and multi-task learning. Through extensive experiments on real-world\ndatasets, we demonstrate that MKR achieves substantial gains in movie, book,\nmusic, and news recommendation, over state-of-the-art baselines. MKR is also\nshown to be able to maintain a decent performance even if user-item\ninteractions are sparse.\n",
"title": "Multi-Task Feature Learning for Knowledge Graph Enhanced Recommendation"
}
| null | null | null | null | true | null |
13327
| null |
Default
| null | null |
null |
{
"abstract": " Graph-based semi-supervised learning is one of the most popular methods in\nmachine learning. Some of its theoretical properties such as bounds for the\ngeneralization error and the convergence of the graph Laplacian regularizer\nhave been studied in computer science and statistics literatures. However, a\nfundamental statistical property, the consistency of the estimator from this\nmethod has not been proved. In this article, we study the consistency problem\nunder a non-parametric framework. We prove the consistency of graph-based\nlearning in the case that the estimated scores are enforced to be equal to the\nobserved responses for the labeled data. The sample sizes of both labeled and\nunlabeled data are allowed to grow in this result. When the estimated scores\nare not required to be equal to the observed responses, a tuning parameter is\nused to balance the loss function and the graph Laplacian regularizer. We give\na counterexample demonstrating that the estimator for this case can be\ninconsistent. The theoretical findings are supported by numerical studies.\n",
"title": "On Consistency of Graph-based Semi-supervised Learning"
}
| null | null |
[
"Statistics"
] | null | true | null |
13328
| null |
Validated
| null | null |
null |
{
"abstract": " List-wise learning to rank methods are considered to be the state-of-the-art.\nOne of the major problems with these methods is that the ambiguous nature of\nrelevance labels in learning to rank data is ignored. Ambiguity of relevance\nlabels refers to the phenomenon that multiple documents may be assigned the\nsame relevance label for a given query, so that no preference order should be\nlearned for those documents. In this paper we propose a novel sampling\ntechnique for computing a list-wise loss that can take into account this\nambiguity. We show the effectiveness of the proposed method by training a\n3-layer deep neural network. We compare our new loss function to two strong\nbaselines: ListNet and ListMLE. We show that our method generalizes better and\nsignificantly outperforms other methods on the validation and test sets.\n",
"title": "Modeling Label Ambiguity for Neural List-Wise Learning to Rank"
}
| null | null | null | null | true | null |
13329
| null |
Default
| null | null |
null |
{
"abstract": " We study the Whitehead torsions of inertial h-cobordisms, and identify\nvarious types representing a nested sequence of subsets of the Whitehead group.\nA number of examples are given to show that these subsets are all different in\ngeneral.\n",
"title": "Whitehead torsion of inertial h-cobordisms"
}
| null | null | null | null | true | null |
13330
| null |
Default
| null | null |
null |
{
"abstract": " As people rely on social media as their primary sources of news, the spread\nof misinformation has become a significant concern. In this large-scale study\nof news in social media we analyze eleven million posts and investigate\npropagation behavior of users that directly interact with news accounts\nidentified as spreading trusted versus malicious content. Unlike previous work,\nwhich looks at specific rumors, topics, or events, we consider all content\npropagated by various news sources. Moreover, we analyze and contrast\npopulation versus sub-population behaviour (by demographics) when spreading\nmisinformation, and distinguish between two types of propagation, i.e., direct\nretweets and mentions. Our evaluation examines how evenly, how many, how\nquickly, and which users propagate content from various types of news sources\non Twitter.\nOur analysis has identified several key differences in propagation behavior\nfrom trusted versus suspicious news sources. These include high inequity in the\ndiffusion rate based on the source of disinformation, with a small group of\nhighly active users responsible for the majority of disinformation spread\noverall and within each demographic. Analysis by demographics showed that users\nwith lower annual income and education share more from disinformation sources\ncompared to their counterparts. News content is shared significantly more\nquickly from trusted, conspiracy, and disinformation sources compared to\nclickbait and propaganda. Older users propagate news from trusted sources more\nquickly than younger users, but they share from suspicious sources after longer\ndelays. Finally, users who interact with clickbait and conspiracy sources are\nlikely to share from propaganda accounts, but not the other way around.\n",
"title": "Propagation from Deceptive News Sources: Who Shares, How Much, How Evenly, and How Quickly?"
}
| null | null | null | null | true | null |
13331
| null |
Default
| null | null |
null |
{
"abstract": " Quantum sensors with solid state electron spins have attracted considerable\ninterest due to their nanoscale spatial resolution.A critical requirement is to\nsuppress the environment noise of the solid state spin sensor.Here we\ndemonstrate a nanoscale thermometer based on silicon carbide (SiC) electron\nspins.We experimentally demonstrate that the performance of the spin sensor is\nrobust against dephasing due to a self protected machenism. The SiC thermometry\nmay provide a promising platform for sensing in a noisy environment ,e.g.\nbiological system sensing.\n",
"title": "Self-protected nanoscale thermometry based on spin defects in silicon carbide"
}
| null | null |
[
"Physics"
] | null | true | null |
13332
| null |
Validated
| null | null |
null |
{
"abstract": " Advanced motor skills are essential for robots to physically coexist with\nhumans. Much research on robot dynamics and control has achieved success on\nhyper robot motor capabilities, but mostly through heavily case-specific\nengineering. Meanwhile, in terms of robot acquiring skills in a ubiquitous\nmanner, robot learning from human demonstration (LfD) has achieved great\nprogress, but still has limitations handling dynamic skills and compound\nactions. In this paper, we present a composite learning scheme which goes\nbeyond LfD and integrates robot learning from human definition, demonstration,\nand evaluation. The method tackles advanced motor skills that require dynamic\ntime-critical maneuver, complex contact control, and handling partly soft\npartly rigid objects. We also introduce the \"nunchaku flipping challenge\", an\nextreme test that puts hard requirements to all these three aspects. Continued\nfrom our previous presentations, this paper introduces the latest update of the\ncomposite learning scheme and the physical success of the nunchaku flipping\nchallenge.\n",
"title": "Robot Composite Learning and the Nunchaku Flipping Challenge"
}
| null | null | null | null | true | null |
13333
| null |
Default
| null | null |
null |
{
"abstract": " Querying graph databases has recently received much attention. We propose a\nnew approach to this problem, which balances competing goals of expressive\npower, language clarity and computational complexity. A distinctive feature of\nour approach is the ability to express properties of minimal (e.g. shortest)\nand maximal (e.g. most valuable) paths satisfying given criteria. To express\ncomplex properties in a modular way, we introduce labelling-generating\nontologies. The resulting formalism is computationally attractive -- queries\ncan be answered in non-deterministic logarithmic space in the size of the\ndatabase.\n",
"title": "Querying Best Paths in Graph Databases"
}
| null | null | null | null | true | null |
13334
| null |
Default
| null | null |
null |
{
"abstract": " We show a communication complexity lower bound for finding a correlated\nequilibrium of a two-player game. More precisely, we define a two-player $N\n\\times N$ game called the 2-cycle game and show that the randomized\ncommunication complexity of finding a 1/poly($N$)-approximate correlated\nequilibrium of the 2-cycle game is $\\Omega(N)$. For small approximation values,\nthis answers an open question of Babichenko and Rubinstein (STOC 2017). Our\nlower bound is obtained via a direct reduction from the unique set disjointness\nproblem.\n",
"title": "Communication Complexity of Correlated Equilibrium in Two-Player Games"
}
| null | null | null | null | true | null |
13335
| null |
Default
| null | null |
null |
{
"abstract": " We consider graph Turing machines, a model of parallel computation on a\ngraph, in which each vertex is only capable of performing one of a finite\nnumber of operations. This model of computation is a natural generalization of\nseveral well-studied notions of computation, including ordinary Turing\nmachines, cellular automata, and parallel graph dynamical systems. We analyze\nthe power of computations that can take place in this model, both in terms of\nthe degrees of computability of the functions that can be computed, and the\ntime and space resources needed to carry out these computations. We further\nshow that properties of the underlying graph have significant consequences for\nthe power of computation thereby obtained. In particular, we show that every\narithmetically definable set can be computed by a graph Turing machine in\nconstant time, and that every computably enumerable Turing degree can be\ncomputed in constant time and linear space by a graph Turing machine whose\nunderlying graph has finite degree.\n",
"title": "On the computability of graph Turing machines"
}
| null | null | null | null | true | null |
13336
| null |
Default
| null | null |
null |
{
"abstract": " Many online social networks allow directed edges: Alice can unilaterally add\nan \"edge\" to Bob, typically indicating interest in Bob or Bob's content,\nwithout Bob's permission or reciprocation. In directed social networks we\nobserve the rise of two distinctive classes of users: celebrities who accrue\nunreciprocated incoming links, and follow spammers, who generate unreciprocated\noutgoing links. Identifying users in these two classes is important for abuse\ndetection, user and content ranking, privacy choices, and other social network\nfeatures.\nIn this paper we develop SCRank, an iterative algorithm to identify such\nusers. We analyze SCRank both theoretically and experimentally. The\nspammer-celebrity definition is not amenable to analysis using standard power\niteration, so we develop a novel potential function argument to show\nconvergence to an approximate equilibrium point for a class of algorithms\nincluding SCRank. We then use experimental evaluation on a real global-scale\nsocial network and on synthetically generated graphs to observe that the\nalgorithm converges quickly and consistently. Using synthetic data with\nbuilt-in ground truth, we also experimentally show that the algorithm provides\na good approximation to planted celebrities and spammers.\n",
"title": "SCRank: Spammer and Celebrity Ranking in Directed Social Networks"
}
| null | null | null | null | true | null |
13337
| null |
Default
| null | null |
null |
{
"abstract": " About six years ago, semitoric systems on 4-dimensional manifolds were\nclassified by Pelayo & Vu Ngoc by means of five invariants. A standard example\nof such a system is the coupled spin-oscillator on $\\mathbb{S}^2 \\times\n\\mathbb{R}^2$. Calculations of three of the five semitoric invariants of this\nsystem (namely the number of focus-focus singularities, the generalised\nsemitoric polygon, and the height invariant) already appeared in the\nliterature, but the so-called twisting index was not yet computed and, of the\nso-called Taylor series invariant, only the linear terms were known.\nIn the present paper, we complete the list of invariants for the coupled\nspin-oscillator by calculating higher order terms of the Taylor series\ninvariant and by computing the twisting index. Moreover, we prove that the\nTaylor series invariant has certain symmetry properties that make the even\npowers in one of the variables vanish and allow us to show superintegrability\nof the coupled spin-oscillator on the zero energy level.\n",
"title": "Taylor series and twisting-index invariants of coupled spin-oscillators"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13338
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the $K$-User Multiple-Input-Single-Output (MISO) Broadcast\nChannel (BC) where the transmitter, equipped with $M$ antennas, serves $K$\nusers, with $K \\leq M$. The transmitter has access to a partial channel state\ninformation of the users. This is modelled by letting the variance of the\nChannel State Information at the Transmitter (CSIT) error of user $i$ scale as\n$O(P^{-\\alpha_i}$) for the Signal-to-Noise Ratio (SNR) $P$ and some constant\n$\\alpha_i \\geq 0$. In this work we derive the optimal Degrees-of-Freedom (DoF)\nregion in such setting and we show that Rate-Splitting (RS) is the key scheme\nto achieve such a region.\n",
"title": "Optimal DoF region of the K-User MISO BC with Partial CSIT"
}
| null | null | null | null | true | null |
13339
| null |
Default
| null | null |
null |
{
"abstract": " We simulate complex fluids by means of an on-the-fly coupling of the bulk\nrheology to the underlying microstructure dynamics. In particular, a\nmacroscopic continuum model of polymeric fluids is constructed without a\npre-specified constitutive relation, but instead it is actively learned from\nmesoscopic simulations where the dynamics of polymer chains is explicitly\ncomputed. To couple the macroscopic rheology of polymeric fluids and the\nmicroscale dynamics of polymer chains, the continuum approach (based on the\nfinite volume method) provides the transient flow field as inputs for the\n(mesoscopic) dissipative particle dynamics (DPD), and in turn DPD returns an\neffective constitutive relation to close the continuum equations. In this\nmultiscale modeling procedure, we employ an active learning strategy based on\nGaussian process regression (GPR) to minimize the number of expensive DPD\nsimulations, where adaptively selected DPD simulations are performed only as\nnecessary. Numerical experiments are carried out for flow past a circular\ncylinder of a non-Newtonian fluid, modeled at the mesoscopic level by\nbead-spring chains. The results show that only five DPD simulations are\nrequired to achieve an effective closure of the continuum equations at Reynolds\nnumber Re=10. Furthermore, when Re is increased to 100, only one additional DPD\nsimulation is required for constructing an extended GPR-informed model closure.\nCompared to traditional message-passing multiscale approaches, applying an\nactive learning scheme to multiscale modeling of non-Newtonian fluids can\nsignificantly increase the computational efficiency. Although the method\ndemonstrated here obtains only a local viscosity from the mesoscopic model, it\ncan be extended to other multiscale models of complex fluids whose\nmacro-rheology is unknown.\n",
"title": "Active learning of constitutive relation from mesoscopic dynamics for macroscopic modeling of non-Newtonian flows"
}
| null | null | null | null | true | null |
13340
| null |
Default
| null | null |
null |
{
"abstract": " Motile organisms often use finite spatial perception of their surroundings to\nnavigate and search their habitats. Yet standard models of search are usually\nbased on purely local sensory information. To model how a finite perceptual\nhorizon affects ecological search, we propose a framework for optimal\nnavigation that combines concepts from random walks and optimal control theory.\nWe show that, while local strategies are optimal on asymptotically long and\nshort search times, finite perception yields faster convergence and increased\nsearch efficiency over transient time scales relevant in biological systems.\nThe benefit of the finite horizon can be maintained by the searchers tuning\ntheir response sensitivity to the length scale of the stimulant in the\nenvironment, and is enhanced when the agents interact as a result of increased\nconsensus within subpopulations. Our framework sheds light on the role of\nspatial perception and transients in search movement and collective sensing of\nthe environment.\n",
"title": "Collective search with finite perception: transient dynamics and search efficiency"
}
| null | null | null | null | true | null |
13341
| null |
Default
| null | null |
null |
{
"abstract": " We study the asymptotic behaviour of the twisted first moment of central\n$L$-values associated to cusp forms in weight aspect on average. Our estimate\nof the error term allows extending the logarithmic length of mollifier $\\Delta$\nup to 2. The best previously known result, due to Iwaniec and Sarnak, was\n$\\Delta<1$. The proof is based on a representation formula for the error in\nterms of Legendre polynomials.\n",
"title": "The first moment of cusp form L-functions in weight aspect on average"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13342
| null |
Validated
| null | null |
null |
{
"abstract": " Boltzmann sampling is commonly used to uniformly sample objects of a\nparticular size from large combinatorial sets. For this technique to be\neffective, one needs to prove that (1) the sampling procedure is efficient and\n(2) objects of the desired size are generated with sufficiently high\nprobability. We use this approach to give a provably efficient sampling\nalgorithm for a class of weighted integer partitions related to Bose-Einstein\ncondensation from statistical physics. Our sampling algorithm is a\nprobabilistic interpretation of the ordinary generating function for these\nobjects, derived from the symbolic method of analytic combinatorics. Using the\nKhintchine-Meinardus probabilistic method to bound the rejection rate of our\nBoltzmann sampler through singularity analysis of Dirichlet generating\nfunctions, we offer an alternative approach to analyze Boltzmann samplers for\nobjects with multiplicative structure.\n",
"title": "Analyzing Boltzmann Samplers for Bose-Einstein Condensates with Dirichlet Generating Functions"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13343
| null |
Validated
| null | null |
null |
{
"abstract": " We classify the ergodic invariant random subgroups of block-diagonal limits\nof symmetric groups in the cases when the groups are simple and the associated\ndimension groups have finite dimensional state spaces. These block-diagonal\nlimits arise as the transformation groups (full groups) of Bratteli diagrams\nthat preserve the cofinality of infinite paths in the diagram. Given a simple\nfull group $G$ admitting only a finite number of ergodic measures on the\npath-space $X$ of the associated Bratteli digram, we prove that every non-Dirac\nergodic invariant random subgroup of $G$ arises as the stabilizer distribution\nof the diagonal action on $X^n$ for some $n\\geq 1$. As a corollary, we\nestablish that every group character $\\chi$ of $G$ has the form $\\chi(g) =\nProb(g\\in K)$, where $K$ is a conjugation-invariant random subgroup of $G$.\n",
"title": "On Invariant Random Subgroups of Block-Diagonal Limits of Symmetric Groups"
}
| null | null | null | null | true | null |
13344
| null |
Default
| null | null |
null |
{
"abstract": " Sparse Subspace Clustering (SSC) is a popular unsupervised machine learning\nmethod for clustering data lying close to an unknown union of low-dimensional\nlinear subspaces; a problem with numerous applications in pattern recognition\nand computer vision. Even though the behavior of SSC for complete data is by\nnow well-understood, little is known about its theoretical properties when\napplied to data with missing entries. In this paper we give theoretical\nguarantees for SSC with incomplete data, and analytically establish that\nprojecting the zero-filled data onto the observation pattern of the point being\nexpressed leads to a substantial improvement in performance. The main insight\nthat stems from our analysis is that even though the projection induces\nadditional missing entries, this is counterbalanced by the fact that the\nprojected and zero-filled data are in effect incomplete points associated with\nthe union of the corresponding projected subspaces, with respect to which the\npoint being expressed is complete. The significance of this phenomenon\npotentially extends to the entire class of self-expressive methods.\n",
"title": "Theoretical Analysis of Sparse Subspace Clustering with Missing Entries"
}
| null | null | null | null | true | null |
13345
| null |
Default
| null | null |
null |
{
"abstract": " We consider a certain type of geometric properties of Banach spaces, which\nincludes for instance octahedrality, almost squareness, lushness and the\nDaugavet property. For this type of properties, we obtain a general reduction\ntheorem, which, roughly speaking, states the following: if the property in\nquestion is stable under certain finite absolute sums (for example finite\n$\\ell^p$-sums), then it is also stable under the formation of corresponding\nKöthe-Bochner spaces (for example $L^p$-Bochner spaces). From this general\ntheorem, we obtain as corollaries a number of new results as well as some\nalternative proofs of already known results concerning octahedral and almost\nsquare spaces and their relatives, diameter-two-properties, lush spaces and\nother classes.\n",
"title": "On certain geometric properties in Banach spaces of vector-valued functions"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13346
| null |
Validated
| null | null |
null |
{
"abstract": " Languages shared by people differ in different regions based on their\naccents, pronunciation and word usages. In this era sharing of language takes\nplace mainly through social media and blogs. Every second swing of such a micro\nposts exist which induces the need of processing those micro posts, in-order to\nextract knowledge out of it. Knowledge extraction differs with respect to the\napplication in which the research on cognitive science fed the necessities for\nthe same. This work further moves forward such a research by extracting\nsemantic information of streaming and batch data in applications like Named\nEntity Recognition and Author Profiling. In the case of Named Entity\nRecognition context of a single micro post has been utilized and context that\nlies in the pool of micro posts were utilized to identify the sociolect aspects\nof the author of those micro posts. In this work Conditional Random Field has\nbeen utilized to do the entity recognition and a novel approach has been\nproposed to find the sociolect aspects of the author (Gender, Age group).\n",
"title": "Social Media Analysis based on Semanticity of Streaming and Batch Data"
}
| null | null | null | null | true | null |
13347
| null |
Default
| null | null |
null |
{
"abstract": " We study the phase diagram and edge states of a two-dimensional p-wave\nsuperconductor with long-range hopping and pairing amplitudes. New topological\nphases and quasiparticles different from the usual short-range model are\nobtained. When both hopping and pairing terms decay with the same exponent, one\nof the topological chiral phases with propagating Majorana edge states gets\nsignificantly enhanced by long-range couplings. On the other hand, when the\nlong-range pairing amplitude decays more slowly than the hopping, we discover\nnew topological phases where propagating Majorana fermions at each edge pair\nnonlocally and become gapped even in the thermodynamic limit. Remarkably, these\nnonlocal edge states are still robust, remain separated from the bulk, and are\nlocalized at both edges at the same time. The inclusion of long-range effects\nis potentially applicable to recent experiments with magnetic impurities and\nislands in 2D superconductors.\n",
"title": "Chiral Topological Superconductors Enhanced by Long-Range Interactions"
}
| null | null | null | null | true | null |
13348
| null |
Default
| null | null |
null |
{
"abstract": " We overview the logic of Bunched Implications (BI) and Separation Logic (SL)\nfrom a perspective inspired by Hiroakira Ono's algebraic approach to\nsubstructural logics. We propose generalized BI algebras (GBI-algebras) as a\ncommon framework for algebras arising via \"declarative resource reading\",\nintuitionistic generalizations of relation algebras and arrow logics and the\ndistributive Lambek calculus with intuitionistic implication. Apart from\nexisting models of BI (in particular, heap models and effect algebras), we also\ncover models arising from weakening relations, formal languages or more\nfine-grained treatment of labelled trees and semistructured data. After briefly\ndiscussing the lattice of subvarieties of GBI, we present a suitable duality\nfor GBI along the lines of Esakia and Priestley and an algebraic proof of cut\nelimination in the setting of residuated frames of Galatos and Jipsen. We also\nshow how the algebraic approach allows generic results on decidability, both\npositive and negative ones. In the final part of the paper, we gently introduce\nthe substructural audience to some theory behind state-of-art tools,\nculminating with an algebraic and proof-theoretic presentation of\n(bi-)abduction.\n",
"title": "An Algebraic Glimpse at Bunched Implications and Separation Logic"
}
| null | null | null | null | true | null |
13349
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\,\\Xi\\,$ be the crown domain associated with a non-compact irreducible\nhermitian symmetric space $\\,G/K$. We give an explicit description of the\nunique $\\,G$-invariant adapted hyper-Kähler structure on $\\,\\Xi$,$\\\n$i.$\\,$e.$\\ $compatible with the adapted complex structure $\\,J_{ad}\\,$ and\nwith the $\\,G$-invariant Kähler structure of $\\,G/K$. We also compute\ninvariant potentials of the involved Kähler metrics and the associated moment\nmaps.\n",
"title": "The adapted hyper-Kähler structure on the crown domain"
}
| null | null | null | null | true | null |
13350
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we consider the continuous mathematical model of tumour growth\nand invasion based on the model introduced by Anderson, Chaplain et al.\n\\cite{Anderson&Chaplain2000}, for the case of one space dimension. The model\nconsists of a system of three coupled nonlinear reaction-diffusion-taxis\npartial differential equations describing the interactions between cancer\ncells, the matrix degrading enzyme and the tissue. For this model under certain\nconditions on the model parameters we obtain the exact analytical solutions in\nterms of traveling wave variables. These solutions are smooth positive definite\nfunctions whose profiles agree with those obtained from numerical computations\n\\cite{Chaplain&Lolas2006} for not very large time intervals.\n",
"title": "Exact traveling wave solutions of 1D model of cancer invasion"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
13351
| null |
Validated
| null | null |
null |
{
"abstract": " Let $Q$ be a finite quiver without loops and $\\mathcal{Q}_{\\alpha}$ be the\nLusztig category for any dimension vector $\\alpha$. The purpose of this paper\nis to prove that all Frobenius eigenvalues of the $i$-th cohomology\n$\\mathcal{H}^i(\\mathcal{L})|_x$ for a simple perverse sheaf $\\mathcal{L}\\in\n\\mathcal{Q}_{\\alpha}$ and $x\\in\n\\mathbb{E}_{\\alpha}^{F^n}=\\mathbb{E}_{\\alpha}(\\mathbb{F}_{q^n})$ are equal to\n$(\\sqrt{q^n})^{i}$ as a conjecture given by Schiffmann (\\cite{Schiffmann2}). As\nan application, we prove the existence of a class of Hall polynomials.\n",
"title": "On purity theorem of Lusztig's perverse sheaves"
}
| null | null | null | null | true | null |
13352
| null |
Default
| null | null |
null |
{
"abstract": " The traditional abstract domain framework for imperative programs suffers\nfrom several shortcomings; in particular it does not allow precise symbolic\nabstractions. To solve these problems, we propose a new abstract interpretation\nframework, based on symbolic expressions used both as an abstraction of the\nprogram, and as the input analyzed by abstract domains. We demonstrate new\napplications of the frame- work: an abstract domain that efficiently propagates\nconstraints across the whole program; a new formalization of functor domains as\napproximate translation, which allows the production of approximate programs,\non which we can perform classical symbolic techniques. We used these to build a\ncomplete analyzer for embedded C programs, that demonstrates the practical\napplicability of the framework.\n",
"title": "Abstract Interpretation using a Language of Symbolic Approximation"
}
| null | null | null | null | true | null |
13353
| null |
Default
| null | null |
null |
{
"abstract": " We study the integral transform which appeared in a different form in\nAkhiezer's textbook \"Lectures on Integral Transforms\".\n",
"title": "The Integral Transform of N.I.Akhiezer"
}
| null | null | null | null | true | null |
13354
| null |
Default
| null | null |
null |
{
"abstract": " Quantitative CBA is a postprocessing algorithm for association rule\nclassification algorithm CBA (Liu et al, 1998). QCBA uses original,\nundiscretized numerical attributes to optimize the discovered association\nrules, refining the boundaries of literals in the antecedent of the rules\nproduced by CBA. Some rules as well as literals from the rules can consequently\nbe removed, which makes the resulting classifier smaller. One-rule\nclassification and crisp rules make CBA classification models possibly most\ncomprehensible among all association rule classification algorithms. These\nviable properties are retained by QCBA. The postprocessing is conceptually\nfast, because it is performed on a relatively small number of rules that passed\ndata coverage pruning in CBA. Benchmark of our QCBA approach on 22 UCI datasets\nshows average 53% decrease in the total size of the model as measured by the\ntotal number of conditions in all rules. Model accuracy remains on the same\nlevel as for CBA.\n",
"title": "Quantitative CBA: Small and Comprehensible Association Rule Classification Models"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
13355
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the quermassintegral preserving flow of closed \\emph{h-convex}\nhypersurfaces in hyperbolic space with the speed given by any positive power of\na smooth symmetric, strictly increasing, and homogeneous of degree one function\n$f$ of the principal curvatures which is inverse concave and has dual $f_*$\napproaching zero on the boundary of the positive cone. We prove that if the\ninitial hypersurface is \\emph{h-convex}, then the solution of the flow becomes\nstrictly \\emph{h-convex} for $t>0$, the flow exists for all time and converges\nto a geodesic sphere exponentially in the smooth topology.\n",
"title": "Quermassintegral preserving curvature flow in Hyperbolic space"
}
| null | null | null | null | true | null |
13356
| null |
Default
| null | null |
null |
{
"abstract": " In this article we consider cloaking for a quasi-linear elliptic partial\ndifferential equation of divergence type defined on a bounded domain in\n$\\mathbb{R}^N$ for $N=2,3$. We show that a perfect cloak can be obtained via a\nsingular change of variables scheme and an approximate cloak can be achieved\nvia a regular change of variables scheme. These approximate cloaks though\nnon-degenerate are anisotropic. We also show, within the framework of\nhomogenization, that it is possible to get isotropic regular approximate\ncloaks. This work generalizes to quasi-linear settings previous work on\ncloaking in the context of Electrical Impedance Tomography for the conductivity\nequation.\n",
"title": "Cloaking for a quasi-linear elliptic partial differential equation"
}
| null | null | null | null | true | null |
13357
| null |
Default
| null | null |
null |
{
"abstract": " We consider a novel stochastic multi-armed bandit problem called {\\em good\narm identification} (GAI), where a good arm is defined as an arm with expected\nreward greater than or equal to a given threshold. GAI is a pure-exploration\nproblem that a single agent repeats a process of outputting an arm as soon as\nit is identified as a good one before confirming the other arms are actually\nnot good. The objective of GAI is to minimize the number of samples for each\nprocess. We find that GAI faces a new kind of dilemma, the {\\em\nexploration-exploitation dilemma of confidence}, which is different difficulty\nfrom the best arm identification. As a result, an efficient design of\nalgorithms for GAI is quite different from that for the best arm\nidentification. We derive a lower bound on the sample complexity of GAI that is\ntight up to the logarithmic factor $\\mathrm{O}(\\log \\frac{1}{\\delta})$ for\nacceptance error rate $\\delta$. We also develop an algorithm whose sample\ncomplexity almost matches the lower bound. We also confirm experimentally that\nour proposed algorithm outperforms naive algorithms in synthetic settings based\non a conventional bandit problem and clinical trial researches for rheumatoid\narthritis.\n",
"title": "Good Arm Identification via Bandit Feedback"
}
| null | null |
[
"Statistics"
] | null | true | null |
13358
| null |
Validated
| null | null |
null |
{
"abstract": " Clinical measurements collected over time are naturally represented as\nmultivariate time series (MTS), which often contain missing data. An\nautoencoder can learn low dimensional vectorial representations of MTS that\npreserve important data characteristics, but cannot deal explicitly with\nmissing data. In this work, we propose a new framework that combines an\nautoencoder with the Time series Cluster Kernel (TCK), a kernel that accounts\nfor missingness patterns in MTS. Via kernel alignment, we incorporate TCK in\nthe autoencoder to improve the learned representations in presence of missing\ndata. We consider a classification problem of MTS with missing values,\nrepresenting blood samples of patients with surgical site infection. With our\napproach, rather than with a standard autoencoder, we learn representations in\nlow dimensions that can be classified better.\n",
"title": "Learning compressed representations of blood samples time series with missing data"
}
| null | null | null | null | true | null |
13359
| null |
Default
| null | null |
null |
{
"abstract": " In fairly elementary terms this paper presents how the theory of preordered\nfuzzy sets, more precisely quantale-valued preorders on quantale-valued fuzzy\nsets, is established under the guidance of enriched category theory. Motivated\nby several key results from the theory of quantaloid-enriched categories, this\npaper develops all needed ingredients purely in order-theoretic languages for\nthe readership of fuzzy set theorists, with particular attention paid to fuzzy\nGalois connections between preordered fuzzy sets.\n",
"title": "Fuzzy Galois connections on fuzzy sets"
}
| null | null | null | null | true | null |
13360
| null |
Default
| null | null |
null |
{
"abstract": " We describe an embedding of the QWIRE quantum circuit language in the Coq\nproof assistant. This allows programmers to write quantum circuits using\nhigh-level abstractions and to prove properties of those circuits using Coq's\ntheorem proving features. The implementation uses higher-order abstract syntax\nto represent variable binding and provides a type-checking algorithm for linear\nwire types, ensuring that quantum circuits are well-formed. We formalize a\ndenotational semantics that interprets QWIRE circuits as superoperators on\ndensity matrices, and prove the correctness of some simple quantum programs.\n",
"title": "QWIRE Practice: Formal Verification of Quantum Circuits in Coq"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13361
| null |
Validated
| null | null |
null |
{
"abstract": " We study the problem of learning classifiers with a fairness constraint, with\nthree main contributions towards the goal of quantifying the problem's inherent\ntradeoffs. First, we relate two existing fairness measures to cost-sensitive\nrisks. Second, we show that for cost-sensitive classification and fairness\nmeasures, the optimal classifier is an instance-dependent thresholding of the\nclass-probability function. Third, we show how the tradeoff between accuracy\nand fairness is determined by the alignment between the class-probabilities for\nthe target and sensitive features. Underpinning our analysis is a general\nframework that casts the problem of learning with a fairness requirement as one\nof minimising the difference of two statistical risks.\n",
"title": "The cost of fairness in classification"
}
| null | null | null | null | true | null |
13362
| null |
Default
| null | null |
null |
{
"abstract": " We study the algebraic and analytic structure of Feynman integrals by\nproposing an operation that maps an integral into pairs of integrals obtained\nfrom a master integrand and a corresponding master contour. This operation is a\ncoaction. It reduces to the known coaction on multiple polylogarithms, but\napplies more generally, e.g. to hypergeometric functions. The coaction also\napplies to generic one-loop Feynman integrals with any configuration of\ninternal and external masses, and in dimensional regularization. In this case,\nwe demonstrate that it can be given a diagrammatic representation purely in\nterms of operations on graphs, namely contractions and cuts of edges. The\ncoaction gives direct access to (iterated) discontinuities of Feynman integrals\nand facilitates a straightforward derivation of the differential equations they\nadmit. In particular, the differential equations for any one-loop integral are\ndetermined by the diagrammatic coaction using limited information about their\nmaximal, next-to-maximal, and next-to-next-to-maximal cuts.\n",
"title": "The algebraic structure of cut Feynman integrals and the diagrammatic coaction"
}
| null | null | null | null | true | null |
13363
| null |
Default
| null | null |
null |
{
"abstract": " Geometric phases are well known to be noise-resilient in quantum\nevolutions/operations. Holonomic quantum gates provide us with a robust way\ntowards universal quantum computation, as these quantum gates are actually\ninduced by nonabelian geometric phases. Here we propose and elaborate how to\nefficiently implement universal nonadiabatic holonomic quantum gates on simpler\nsuperconducting circuits, with a single transmon serving as a qubit. In our\nproposal, an arbitrary single-qubit holonomic gate can be realized in a\nsingle-loop scenario, by varying the amplitudes and phase difference of two\nmicrowave fields resonantly coupled to a transmon, while nontrivial two-qubit\nholonomic gates may be generated with a transmission-line resonator being\nsimultaneously coupled to the two target transmons in an effective resonant\nway. Moreover, our scenario may readily be scaled up to a two-dimensional\nlattice configuration, which is able to support large scalable quantum\ncomputation, paving the way for practically implementing universal nonadiabatic\nholonomic quantum computation with superconducting circuits.\n",
"title": "Implementing universal nonadiabatic holonomic quantum gates with transmons"
}
| null | null | null | null | true | null |
13364
| null |
Default
| null | null |
null |
{
"abstract": " Segmented aperture telescopes require an alignment procedure with successive\nsteps from coarse alignment to monitoring process in order to provide very high\noptical quality images for stringent science operations such as exoplanet\nimaging. The final step, referred to as fine phasing, calls for a high\nsensitivity wavefront sensing and control system in a diffraction-limited\nregime to achieve segment alignment with nanometric accuracy. In this context,\nZernike wavefront sensors represent promising options for such a calibration. A\nconcept called the Zernike unit for segment phasing (ZEUS) was previously\ndeveloped for ground-based applications to operate under seeing-limited images.\nSuch a concept is, however, not suitable for fine cophasing with\ndiffraction-limited images. We revisit ZELDA, a Zernike sensor that was\ndeveloped for the measurement of residual aberrations in exoplanet direct\nimagers, to measure segment piston, tip, and tilt in the diffraction-limited\nregime. We introduce a novel analysis scheme of the sensor signal that relies\non piston, tip, and tilt estimators for each segment, and provide probabilistic\ninsights to predict the success of a closed-loop correction as a function of\nthe initial wavefront error. The sensor unambiguously and simultaneously\nretrieves segment piston and tip-tilt misalignment. Our scheme allows for\ncorrection of these errors in closed-loop operation down to nearly zero\nresiduals in a few iterations. This sensor also shows low sensitivity to\nmisalignment of its parts and high ability for operation with a relatively\nbright natural guide star. Our cophasing sensor relies on existing mask\ntechnologies that make the concept already available for segmented apertures in\nfuture space missions.\n",
"title": "Fine cophasing of segmented aperture telescopes with ZELDA, a Zernike wavefront sensor in the diffraction-limited regime"
}
| null | null |
[
"Physics"
] | null | true | null |
13365
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the problem of active feature acquisition, where we sequentially\nselect the subset of features in order to achieve the maximum prediction\nperformance in the most cost-effective way. In this work, we formulate this\nactive feature acquisition problem as a reinforcement learning problem, and\nprovide a novel framework for jointly learning both the RL agent and the\nclassifier (environment). We also introduce a more systematic way of encoding\nsubsets of features that can properly handle innate challenge with missing\nentries in active feature acquisition problems, that uses the orderless\nLSTM-based set encoding mechanism that readily fits in the joint learning\nframework. We evaluate our model on a carefully designed synthetic dataset for\nthe active feature acquisition as well as several real datasets such as\nelectric health record (EHR) datasets, on which it outperforms all baselines in\nterms of prediction performance as well feature acquisition cost.\n",
"title": "Why Pay More When You Can Pay Less: A Joint Learning Framework for Active Feature Acquisition and Classification"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
13366
| null |
Validated
| null | null |
null |
{
"abstract": " We give an asymptotic formula for the number of biquadratic extensions of the\nrationals of bounded discriminant that fail the Hasse norm principle.\n",
"title": "The Hasse Norm Principle For Biquadratic Extensions"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13367
| null |
Validated
| null | null |
null |
{
"abstract": " We consider an optimal execution problem in which a trader is looking at a\nshort-term price predictive signal while trading. In the case where the trader\nis creating an instantaneous market impact, we show that transactions costs\nresulting from the optimal adaptive strategy are substantially lower than the\ncorresponding costs of the optimal static strategy. Later, we investigate the\ncase where the trader is creating transient market impact. We show that\nstrategies in which the trader is observing the signal a number of times during\nthe trading period, can dramatically reduce the transaction costs and improve\nthe performance of the optimal static strategy. These results answer a question\nwhich was raised by Brigo and Piat [6], by analyzing two cases where adaptive\nstrategies can improve the performance of the execution.\n",
"title": "Static vs Adaptive Strategies for Optimal Execution with Signals"
}
| null | null | null | null | true | null |
13368
| null |
Default
| null | null |
null |
{
"abstract": " We give a description of the weighted Reed-Muller codes over a prime field in\na modular algebra. A description of the homogeneous Reed-Muller codes in the\nsame ambient space is presented for the binary case. A decoding procedure using\nthe Landrock-Manz method is developed.\n",
"title": "New descriptions of the weighted Reed-Muller codes and the homogeneous Reed-Muller codes"
}
| null | null | null | null | true | null |
13369
| null |
Default
| null | null |
null |
{
"abstract": " In the junction $\\Omega$ of several semi-infinite cylindrical waveguides we\nconsider the Dirichlet Laplacian whose continuous spectrum is the ray\n$[\\lambda_\\dagger, +\\infty)$ with a positive cut-off value $\\lambda_\\dagger$.\nWe give two different criteria for the threshold resonance generated by\nnontrivial bounded solutions to the Dirichlet problem for the Helmholtz\nequation $-\\Delta u=\\lambda_\\dagger u$ in $\\Omega$. The first criterion is\nquite simple and is convenient to disprove the existence of bounded solutions.\nThe second criterion is rather involved but can help to detect concrete shapes\nsupporting the resonance. Moreover, the latter distinguishes in a natural way\nbetween stabilizing, i.e., bounded but non-descending solutions and trapped\nmodes with exponential decay at infinity.\n",
"title": "Criteria for the Absence and Existence of Bounded Solutions at the Threshold Frequency in a Junction of Quantum Waveguides"
}
| null | null | null | null | true | null |
13370
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the stochastic optimal control problem for systems with\nunknown dynamics. First, an open-loop deterministic trajectory optimization\nproblem is solved without knowing the explicit form of the dynamical system.\nNext, a Linear Quadratic Gaussian (LQG) controller is designed for the nominal\ntrajectory-dependent linearized system, such that under a small noise\nassumption, the actual states remain close to the optimal trajectory. The\ntrajectory-dependent linearized system is identified using input-output\nexperimental data consisting of the impulse responses of the nominal system. A\ncomputational example is given to illustrate the performance of the proposed\napproach.\n",
"title": "Stochastic Feedback Control of Systems with Unknown Nonlinear Dynamics"
}
| null | null | null | null | true | null |
13371
| null |
Default
| null | null |
null |
{
"abstract": " Models and observations suggest that ice-particle aggregation at and beyond\nthe snowline dominates the earliest stages of planet-formation, which therefore\nis subject to many laboratory studies. However, the pressure-temperature\ngradients in proto-planetary disks mean that the ices are constantly processed,\nundergoing phase changes between different solid phases and the gas phase. Open\nquestions remain as to whether the properties of the icy particles themselves\ndictate collision outcomes and therefore how effectively collision experiments\nreproduce conditions in pro- toplanetary environments. Previous experiments\noften yielded apparently contradictory results on collision outcomes, only\nagreeing in a temperature dependence setting in above $\\approx$ 210 K. By\nexploiting the unique capabilities of the NIMROD neutron scattering instrument,\nwe characterized the bulk and surface structure of icy particles used in\ncollision experiments, and studied how these structures alter as a function of\ntemperature at a constant pressure of around 30 mbar. Our icy grains, formed\nunder liquid nitrogen, undergo changes in the crystalline ice-phase,\nsublimation, sintering and surface pre-melting as they are heated from 103 to\n247 K. An increase in the thickness of the diffuse surface layer from $\\approx$\n10 to $\\approx$ 30 {\\AA} ($\\approx$ 2.5 to 12 bilayers) proves increased\nmolecular mobility at temperatures above $\\approx$ 210 K. As none of the other\nchanges tie-in with the temperature trends in collisional outcomes, we conclude\nthat the surface pre-melting phenomenon plays a key role in collision\nexperiments at these temperatures. Consequently, the pressure-temperature\nenvironment, may have a larger influence on collision outcomes than previously\nthought.\n",
"title": "Micrometer-Sized Water Ice Particles for Planetary Science Experiments: Influence of Surface Structure on Collisional Properties"
}
| null | null | null | null | true | null |
13372
| null |
Default
| null | null |
null |
{
"abstract": " The splashback radius $R_{\\rm sp}$, the apocentric radius of particles on\ntheir first orbit after falling into a dark matter halo, has recently been\nsuggested as a physically motivated halo boundary that separates accreting from\norbiting material. Using the SPARTA code presented in Paper I, we analyze the\norbits of billions of particles in cosmological simulations of structure\nformation and measure $R_{\\rm sp}$ for a large sample of halos that span a mass\nrange from dwarf galaxy to massive cluster halos, reach redshift 8, and include\nWMAP, Planck, and self-similar cosmologies. We analyze the dependence of\n$R_{\\rm sp}/R_{\\rm 200m}$ and $M_{\\rm sp}/M_{\\rm 200m}$ on the mass accretion\nrate $\\Gamma$, halo mass, redshift, and cosmology. The scatter in these\nrelations varies between 0.02 and 0.1 dex. While we confirm the known trend\nthat $R_{\\rm sp}/R_{\\rm 200m}$ decreases with $\\Gamma$, the relationships turn\nout to be more complex than previously thought, demonstrating that $R_{\\rm sp}$\nis an independent definition of the halo boundary that cannot trivially be\nreconstructed from spherical overdensity definitions. We present fitting\nfunctions for $R_{\\rm sp}/R_{\\rm 200m}$ and $M_{\\rm sp}/M_{\\rm 200m}$ as a\nfunction of accretion rate, peak height, and redshift, achieving an accuracy of\n5% or better everywhere in the parameter space explored. We discuss the\nphysical meaning of the distribution of particle apocenters and show that the\npreviously proposed definition of $R_{\\rm sp}$ as the radius of the steepest\nlogarithmic density slope encloses roughly three-quarters of the apocenters.\nFinally, we conclude that no analytical model presented thus far can fully\nexplain our results.\n",
"title": "The splashback radius of halos from particle dynamics. II. Dependence on mass, accretion rate, redshift, and cosmology"
}
| null | null |
[
"Physics"
] | null | true | null |
13373
| null |
Validated
| null | null |
null |
{
"abstract": " Deep learning models (DLMs) are state-of-the-art techniques in speech\nrecognition. However, training good DLMs can be time consuming especially for\nproduction-size models and corpora. Although several parallel training\nalgorithms have been proposed to improve training efficiency, there is no clear\nguidance on which one to choose for the task in hand due to lack of systematic\nand fair comparison among them. In this paper we aim at filling this gap by\ncomparing four popular parallel training algorithms in speech recognition,\nnamely asynchronous stochastic gradient descent (ASGD), blockwise model-update\nfiltering (BMUF), bulk synchronous parallel (BSP) and elastic averaging\nstochastic gradient descent (EASGD), on 1000-hour LibriSpeech corpora using\nfeed-forward deep neural networks (DNNs) and convolutional, long short-term\nmemory, DNNs (CLDNNs). Based on our experiments, we recommend using BMUF as the\ntop choice to train acoustic models since it is most stable, scales well with\nnumber of GPUs, can achieve reproducible results, and in many cases even\noutperforms single-GPU SGD. ASGD can be used as a substitute in some cases.\n",
"title": "Empirical Evaluation of Parallel Training Algorithms on Acoustic Modeling"
}
| null | null | null | null | true | null |
13374
| null |
Default
| null | null |
null |
{
"abstract": " Robotic systems are increasingly relying on distributed feedback controllers\nto tackle complex and latency-prone sensing and decision problems. These\ndemands come at the cost of a growing computational burden and, as a result,\nlarger controller latencies. To maximize robustness to mechanical disturbances\nand achieve high control performance, we emphasize the necessity for executing\ndamping feedback in close proximity to the control plant while allocating\nstiffness feedback in a latency-prone centralized control process.\nAdditionally, series elastic actuators (SEAs) are becoming prevalent in\ntorque-controlled robots during recent years to achieve compliant interactions\nwith environments and humans. However, designing optimal impedance controllers\nand characterizing impedance performance for SEAs with time delays and\nfiltering are still under-explored problems. The presented study addresses the\noptimal controller design problem by devising a critically-damped gain design\nmethod for a class of SEA cascaded control architectures, which is composed of\nouter-impedance and inner-torque feedback loops. Via the proposed controller\ndesign criterion, we adopt frequency-domain methods to thoroughly analyze the\neffects of time delays, filtering and load inertia on SEA impedance\nperformance. These results are further validated through the analysis,\nsimulation, and experimental testing on high-performance actuators and on an\nomnidirectional mobile base.\n",
"title": "Distributed Impedance Control of Latency-Prone Robotic Systems with Series Elastic Actuation"
}
| null | null | null | null | true | null |
13375
| null |
Default
| null | null |
null |
{
"abstract": " We consider solution of stochastic storage problems through regression Monte\nCarlo (RMC) methods. Taking a statistical learning perspective, we develop the\ndynamic emulation algorithm (DEA) that unifies the different existing\napproaches in a single modular template. We then investigate the two central\naspects of regression architecture and experimental design that constitute DEA.\nFor the regression piece, we discuss various non-parametric approaches, in\nparticular introducing the use of Gaussian process regression in the context of\nstochastic storage. For simulation design, we compare the performance of\ntraditional design (grid discretization), against space-filling, and several\nadaptive alternatives. The overall DEA template is illustrated with multiple\nexamples drawing from natural gas storage valuation and optimal control of\nback-up generator in a microgrid.\n",
"title": "Simulation Methods for Stochastic Storage Problems: A Statistical Learning Perspective"
}
| null | null | null | null | true | null |
13376
| null |
Default
| null | null |
null |
{
"abstract": " (This is a general physics level overview article about hidden sectors, and\nhow they motivate searches for long-lived particles. Intended for publication\nin Physics Today.)\nSearches for new physics at the Large Hadron Collider have so far come up\nempty, but we just might not be looking in the right place. Spectacular bursts\nof particles appearing seemingly out of nowhere could shed light on some of\nnature's most profound mysteries.\n",
"title": "Flashes of Hidden Worlds at Colliders"
}
| null | null | null | null | true | null |
13377
| null |
Default
| null | null |
null |
{
"abstract": " Fast, byte-addressable non-volatile memory (NVM) embraces both near-DRAM\nlatency and disk-like persistence, which has generated considerable interests\nto revolutionize system software stack and programming models. However, it is\nless understood how NVM can be combined with managed runtime like Java virtual\nmachine (JVM) to ease persistence management. This paper proposes Espresso, a\nholistic extension to Java and its runtime, to enable Java programmers to\nexploit NVM for persistence management with high performance. Espresso first\nprovides a general persistent heap design called Persistent Java Heap (PJH) to\nmanage persistent data as normal Java objects. The heap is then strengthened\nwith a recoverable mechanism to provide crash consistency for heap metadata. It\nthen provides a new abstraction called Persistent Java Object (PJO) to provide\nan easy-to-use but safe persistent programming model for programmers to persist\napplication data. The evaluation confirms that Espresso significantly\noutperforms state-of-art NVM support for Java (i.e., JPA and PCJ) while being\ncompatible to existing data structures in Java programs.\n",
"title": "Espresso: Brewing Java For More Non-Volatility with Non-volatile Memory"
}
| null | null | null | null | true | null |
13378
| null |
Default
| null | null |
null |
{
"abstract": " Our experience of the world is multimodal - we see objects, hear sounds, feel\ntexture, smell odors, and taste flavors. Modality refers to the way in which\nsomething happens or is experienced and a research problem is characterized as\nmultimodal when it includes multiple such modalities. In order for Artificial\nIntelligence to make progress in understanding the world around us, it needs to\nbe able to interpret such multimodal signals together. Multimodal machine\nlearning aims to build models that can process and relate information from\nmultiple modalities. It is a vibrant multi-disciplinary field of increasing\nimportance and with extraordinary potential. Instead of focusing on specific\nmultimodal applications, this paper surveys the recent advances in multimodal\nmachine learning itself and presents them in a common taxonomy. We go beyond\nthe typical early and late fusion categorization and identify broader\nchallenges that are faced by multimodal machine learning, namely:\nrepresentation, translation, alignment, fusion, and co-learning. This new\ntaxonomy will enable researchers to better understand the state of the field\nand identify directions for future research.\n",
"title": "Multimodal Machine Learning: A Survey and Taxonomy"
}
| null | null | null | null | true | null |
13379
| null |
Default
| null | null |
null |
{
"abstract": " Consider the Tate twist $\\tau \\in H^{0,1}(S^{0,0})$ in the mod 2 cohomology\nof the motivic sphere. After 2-completion, the motivic Adams spectral sequence\nrealizes this element as a map $\\tau \\colon S^{0,-1} \\to S^{0,0}$, with cofiber\n$C\\tau$. We show that this motivic 2-cell complex can be endowed with a unique\n$E_{\\infty}$ ring structure. Moreover, this promotes the known isomorphism\n$\\pi_{\\ast,\\ast} C\\tau \\cong\n\\mathrm{Ext}^{\\ast,\\ast}_{BP_{\\ast}BP}(BP_{\\ast},BP_{\\ast})$ to an isomorphism\nof rings which also preserves higher products.\nWe then consider the closed symmetric monoidal category $({\n}_{C\\tau}\\textbf{Mod}, - \\wedge_{C\\tau} -)$ which lives in the kernel of Betti\nrealization. Given a motivic spectrum $X$, the $C\\tau$-induced spectrum $X\n\\wedge C\\tau$ is usually better behaved and easier to understand than $X$\nitself. We specifically illustrate this concept in the examples of the mod 2\nEilenberg-Maclane spectrum $H\\mathbb{F}_2$, the mod 2 Moore spectrum\n$S^{0,0}/2$ and the connective hermitian $K$-theory spectrum $kq$.\n",
"title": "The Motivic Cofiber of $τ$"
}
| null | null | null | null | true | null |
13380
| null |
Default
| null | null |
null |
{
"abstract": " Reliable diagnosis of depressive disorder is essential for both optimal\ntreatment and prevention of fatal outcomes. In this study, we aimed to\nelucidate the effectiveness of two non-linear measures, Higuchi Fractal\nDimension (HFD) and Sample Entropy (SampEn), in detecting depressive disorders\nwhen applied on EEG. HFD and SampEn of EEG signals were used as features for\nseven machine learning algorithms including Multilayer Perceptron, Logistic\nRegression, Support Vector Machines with the linear and polynomial kernel,\nDecision Tree, Random Forest, and Naive Bayes classifier, discriminating EEG\nbetween healthy control subjects and patients diagnosed with depression. We\nconfirmed earlier observations that both non-linear measures can discriminate\nEEG signals of patients from healthy control subjects. The results suggest that\ngood classification is possible even with a small number of principal\ncomponents. Average accuracy among classifiers ranged from 90.24% to 97.56%.\nAmong the two measures, SampEn had better performance. Using HFD and SampEn and\na variety of machine learning techniques we can accurately discriminate\npatients diagnosed with depression vs controls which can serve as a highly\nsensitive, clinically relevant marker for the diagnosis of depressive\ndisorders.\n",
"title": "EEG machine learning with Higuchi fractal dimension and Sample Entropy as features for successful detection of depression"
}
| null | null | null | null | true | null |
13381
| null |
Default
| null | null |
null |
{
"abstract": " Targeted advertising is meant to improve the efficiency of matching\nadvertisers to their customers. However, targeted advertising can also be\nabused by malicious advertisers to efficiently reach people susceptible to\nfalse stories, stoke grievances, and incite social conflict. Since targeted ads\nare not seen by non-targeted and non-vulnerable people, malicious ads are\nlikely to go unreported and their effects undetected. This work examines a\nspecific case of malicious advertising, exploring the extent to which political\nads from the Russian Intelligence Research Agency (IRA) run prior to 2016 U.S.\nelections exploited Facebook's targeted advertising infrastructure to\nefficiently target ads on divisive or polarizing topics (e.g., immigration,\nrace-based policing) at vulnerable sub-populations. In particular, we do the\nfollowing: (a) We conduct U.S. census-representative surveys to characterize\nhow users with different political ideologies report, approve, and perceive\ntruth in the content of the IRA ads. Our surveys show that many ads are\n\"divisive\": they elicit very different reactions from people belonging to\ndifferent socially salient groups. (b) We characterize how these divisive ads\nare targeted to sub-populations that feel particularly aggrieved by the status\nquo. Our findings support existing calls for greater transparency of content\nand targeting of political ads. (c) We particularly focus on how the Facebook\nad API facilitates such targeting. We show how the enormous amount of personal\ndata Facebook aggregates about users and makes available to advertisers enables\nsuch malicious targeting.\n",
"title": "On Microtargeting Socially Divisive Ads: A Case Study of Russia-Linked Ad Campaigns on Facebook"
}
| null | null | null | null | true | null |
13382
| null |
Default
| null | null |
null |
{
"abstract": " We prove a Gauss-Bonnet formula X(G) = sum_x K(x), where K(x)=(-1)^dim(x)\n(1-X(S(x))) is a curvature of a vertex x with unit sphere S(x) in the\nBarycentric refinement G1 of a simplicial complex G. K(x) is dual to\n(-1)^dim(x) for which Gauss-Bonnet is the definition of Euler characteristic X.\nBecause the connection Laplacian L'=1+A' of G is unimodular, where A' is the\nadjacency matrix of of the connection graph G', the Green function values\ng(x,y) = (1+A')^-1_xy are integers and 1-X(S(x))=g(x,x). Gauss-Bonnet for K^+\nreads therefore as str(g)=X(G), where str is the super trace. As g is a\ntime-discrete heat kernel, this is a cousin to McKean-Singer str(exp(-Lt)) =\nX(G) for the Hodge Laplacian L=dd^* +d^*d which lives on the same Hilbert space\nthan L'. Both formulas hold for an arbitrary finite abstract simplicial complex\nG. Writing V_x(y)= g(x,y) for the Newtonian potential of the connection\nLaplacian, we prove sum_y V_x(y) = K(x), so that by the new Gauss-Bonnet\nformula, the Euler characteristic of G agrees with the total potential\ntheoretic energy sum_x,y g(x,y)=X(G) of G. The curvature K now relates to the\nprobability measure p minimizing the internal energy U(p)=sum_x,y g(x,y) p(x)\np(y) of the complex. Since both the internal energy (here linked to topology)\nand Shannon entropy are natural and unique in classes of functionals, we then\nlook at critical points p the Helmholtz free energy F(p)=(1-T) U(p)-T S(p)\nwhich combines the energy functional U and the entropy functional S(p)=-sum_x\np(x) log(p(x)). As the temperature T changes, we observe bifurcation phenomena.\nAlready for G=K_3 both a saddle node bifurcation and a pitchfork bifurcation\noccurs. The saddle node bifurcation leads to a catastrophe: the function T ->\nF(p(T),T) is discontinuous if p(T) is a free energy minimizer.\n",
"title": "On Helmholtz free energy for finite abstract simplicial complexes"
}
| null | null | null | null | true | null |
13383
| null |
Default
| null | null |
null |
{
"abstract": " Social media are transforming global communication and coordination. The data\nderived from social media can reveal patterns of human behavior at all levels\nand scales of society. Using geolocated Twitter data, we have quantified\ncollective behaviors across multiple scales, ranging from the commutes of\nindividuals, to the daily pulse of 50 major urban areas and global patterns of\nhuman coordination. Human activity and mobility patterns manifest the synchrony\nrequired for contingency of actions between individuals. Urban areas show\nregular cycles of contraction and expansion that resembles heartbeats linked\nprimarily to social rather than natural cycles. Business hours and circadian\nrhythms influence daily cycles of work, recreation, and sleep. Different urban\nareas have characteristic signatures of daily collective activities. The\ndifferences are consistent with a new emergent global synchrony that couples\nbehavior in distant regions across the world. A globally synchronized peak that\nincludes exchange of ideas and information across Europe, Africa, Asia and\nAustralasia. We propose a dynamical model to explain the emergence of global\nsynchrony in the context of increasing global communication and reproduce the\nobserved behavior. The collective patterns we observe show how social\ninteractions lead to interdependence of behavior manifest in the\nsynchronization of communication. The creation and maintenance of temporally\nsensitive social relationships results in the emergence of complexity of the\nlarger scale behavior of the social system.\n",
"title": "Global Patterns of Synchronization in Human Communications"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
13384
| null |
Validated
| null | null |
null |
{
"abstract": " Infectious disease outbreaks recapitulate biology: they emerge from the\nmulti-level interaction of hosts, pathogens, and their shared environment. As a\nresult, predicting when, where, and how far diseases will spread requires a\ncomplex systems approach to modeling. Recent studies have demonstrated that\npredicting different components of outbreaks--e.g., the expected number of\ncases, pace and tempo of cases needing treatment, demand for prophylactic\nequipment, importation probability etc.--is feasible. Therefore, advancing both\nthe science and practice of disease forecasting now requires testing for the\npresence of fundamental limits to outbreak prediction. To investigate the\nquestion of outbreak prediction, we study the information theoretic limits to\nforecasting across a broad set of infectious diseases using permutation entropy\nas a model independent measure of predictability. Studying the predictability\nof a diverse collection of historical outbreaks--including, chlamydia, dengue,\ngonorrhea, hepatitis A, influenza, measles, mumps, polio, and whooping\ncough--we identify a fundamental entropy barrier for infectious disease time\nseries forecasting. However, we find that for most diseases this barrier to\nprediction is often well beyond the time scale of single outbreaks. We also\nfind that the forecast horizon varies by disease and demonstrate that both\nshifting model structures and social network heterogeneity are the most likely\nmechanisms for the observed differences across contagions. Our results\nhighlight the importance of moving beyond time series forecasting, by embracing\ndynamic modeling approaches, and suggest challenges for performing model\nselection across long time series. We further anticipate that our findings will\ncontribute to the rapidly growing field of epidemiological forecasting and may\nrelate more broadly to the predictability of complex adaptive systems.\n",
"title": "On the predictability of infectious disease outbreaks"
}
| null | null | null | null | true | null |
13385
| null |
Default
| null | null |
null |
{
"abstract": " Multi-player Multi-Armed Bandits (MAB) have been extensively studied in the\nliterature, motivated by applications to Cognitive Radio systems. Driven by\nsuch applications as well, we motivate the introduction of several levels of\nfeedback for multi-player MAB algorithms. Most existing work assume that\nsensing information is available to the algorithm. Under this assumption, we\nimprove the state-of-the-art lower bound for the regret of any decentralized\nalgorithms and introduce two algorithms, RandTopM and MCTopM, that are shown to\nempirically outperform existing algorithms. Moreover, we provide strong\ntheoretical guarantees for these algorithms, including a notion of asymptotic\noptimality in terms of the number of selections of bad arms. We then introduce\na promising heuristic, called Selfish, that can operate without sensing\ninformation, which is crucial for emerging applications to Internet of Things\nnetworks. We investigate the empirical performance of this algorithm and\nprovide some first theoretical elements for the understanding of its behavior.\n",
"title": "Multi-Player Bandits Revisited"
}
| null | null | null | null | true | null |
13386
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we consider the estimation of generalized linear models with\ncovariates that are missing completely at random. We propose a model averaging\nestimation method and prove that the corresponding model averaging estimator is\nasymptotically optimal under certain assumptions. Simulaiton results illustrate\nthat this method has better performance than other alternatives under most\nsituations.\n",
"title": "Model Averaging for Generalized Linear Model with Covariates that are Missing completely at Random"
}
| null | null | null | null | true | null |
13387
| null |
Default
| null | null |
null |
{
"abstract": " Light-shining-through-a-wall experiments represent a new experimental\napproach in the search for undiscovered elementary particles not accessible\nwith accelerator based experiments. The next generation of these experiments,\nsuch as ALPS~II, require high finesse, long baseline optical cavities with fast\nlength control. In this paper we report on a length stabilization control loop\nused to keep a 9.2\\,m cavity resonant. It achieves a unity-gain-frequency of\n4\\,kHz and actuates on a mirror with a diameter of 50.8\\,mm. The finesse of\nthis cavity was measured to be 101,304$\\pm$540 for 1064\\,nm light. The\ndifferential cavity length noise between 1064\\,nm and 532\\,nm light was also\nmeasured since 532\\,nm light will be used to sense the length of the\nregeneration cavity. Out-of-loop noise sources and different control strategies\nare discussed, in order to fulfill the length stability requirements for\nALPS~II.\n",
"title": "Demonstration of the length stability requirements for ALPS II with a high finesse 9.2m cavity"
}
| null | null | null | null | true | null |
13388
| null |
Default
| null | null |
null |
{
"abstract": " The recent increase of interest in the graph invariant called tree-depth and\nin its applications in algorithms and logic on graphs led to a natural\nquestion: is there an analogously useful \"depth\" notion also for dense graphs\n(say; one which is stable under graph complementation)? To this end, in a 2012\nconference paper, a new notion of shrub-depth has been introduced, such that it\nis related to the established notion of clique-width in a similar way as\ntree-depth is related to tree-width. Since then shrub-depth has been\nsuccessfully used in several research papers. Here we provide an in-depth\nreview of the definition and basic properties of shrub-depth, and we focus on\nits logical aspects which turned out to be most useful. In particular, we use\nshrub-depth to give a characterization of the lower ${\\omega}$ levels of the\nMSO1 transduction hierarchy of simple graphs.\n",
"title": "Shrub-depth: Capturing Height of Dense Graphs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
13389
| null |
Validated
| null | null |
null |
{
"abstract": " Interpreting neural networks is a crucial and challenging task in machine\nlearning. In this paper, we develop a novel framework for detecting statistical\ninteractions captured by a feedforward multilayer neural network by directly\ninterpreting its learned weights. Depending on the desired interactions, our\nmethod can achieve significantly better or similar interaction detection\nperformance compared to the state-of-the-art without searching an exponential\nsolution space of possible interactions. We obtain this accuracy and efficiency\nby observing that interactions between input features are created by the\nnon-additive effect of nonlinear activation functions, and that interacting\npaths are encoded in weight matrices. We demonstrate the performance of our\nmethod and the importance of discovered interactions via experimental results\non both synthetic datasets and real-world application datasets.\n",
"title": "Detecting Statistical Interactions from Neural Network Weights"
}
| null | null | null | null | true | null |
13390
| null |
Default
| null | null |
null |
{
"abstract": " Intersections are hazardous places. Threats arise from interactions among\npedestrians, bicycles and vehicles, more complicated vehicle trajectories in\nthe absence of lane markings, phases that prevent determining who has the right\nof way, invisible vehicle approaches, vehicle obstructions, and illegal\nmovements. These challenges are not fully addressed by the \"road diet\" and road\nredesign prescribed in Vision Zero plans, nor will they be completely overcome\nby autonomous vehicles with their many sensors and tireless attention to\nsurroundings. Accidents can also occur because drivers, cyclists and\npedestrians do not have the information they need to avoid wrong decisions. In\nthese cases, the missing information can be computed and broadcast by an\nintelligent intersection. The information gives the current full signal phase,\nan estimate of the time when the phase will change, and the occupancy of the\nblind spots of the driver or autonomous vehicle. The paper develops a design of\nthe intelligent intersection, motivated by the analysis of an accident at an\nintersection in Tempe, AZ, between an automated Uber Volvo and a manual Honda\nCRV and culminates in a proposal for an intelligent intersection\ninfrastructure. The intelligent intersection also serves as a software-enabled\nversion of the `protected intersection' design to improve the passage of\ncyclists and pedestrians through an intersection.\n",
"title": "Making intersections safer with I2V communication"
}
| null | null | null | null | true | null |
13391
| null |
Default
| null | null |
null |
{
"abstract": " While convolutional sparse representations enjoy a number of useful\nproperties, they have received limited attention for image reconstruction\nproblems. The present paper compares the performance of block-based and\nconvolutional sparse representations in the removal of Gaussian white noise.\nWhile the usual formulation of the convolutional sparse coding problem is\nslightly inferior to the block-based representations in this problem, the\nperformance of the convolutional form can be boosted beyond that of the\nblock-based form by the inclusion of suitable penalties on the gradients of the\ncoefficient maps.\n",
"title": "Convolutional Sparse Representations with Gradient Penalties"
}
| null | null | null | null | true | null |
13392
| null |
Default
| null | null |
null |
{
"abstract": " MMS observations recently confirmed that crescent-shaped electron velocity\ndistributions in the plane perpendicular to the magnetic field occur in the\nelectron diffusion region near reconnection sites at Earth's magnetopause. In\nthis paper, we re-examine the origin of the crescent-shaped distributions in\nthe light of our new finding that ions and electrons are drifting in opposite\ndirections when displayed in magnetopause boundary-normal coordinates.\nTherefore, ExB drifts cannot cause the crescent shapes. We performed a\nhigh-resolution multi-scale simulation capturing sub-electron skin depth\nscales. The results suggest that the crescent-shaped distributions are caused\nby meandering orbits without necessarily requiring any additional processes\nfound at the magnetopause such as the highly asymmetric magnetopause ambipolar\nelectric field. We use an adiabatic Hamiltonian model of particle motion to\nconfirm that conservation of canonical momentum in the presence of magnetic\nfield gradients causes the formation of crescent shapes without invoking\nasymmetries or the presence of an ExB drift. An important consequence of this\nfinding is that we expect crescent-shaped distributions also to be observed in\nthe magnetotail, a prediction that MMS will soon be able to test.\n",
"title": "On the origin of the crescent-shaped distributions observed by MMS at the magnetopause"
}
| null | null | null | null | true | null |
13393
| null |
Default
| null | null |
null |
{
"abstract": " In an era where big and high-dimensional data is readily available, data\nscientists are inevitably faced with the challenge of reducing this data for\nexpensive downstream computation or analysis. To this end, we present here a\nnew method for reducing high-dimensional big data into a representative point\nset, called projected support points (PSPs). A key ingredient in our method is\nthe so-called sparsity-inducing (SpIn) kernel, which encourages the\npreservation of low-dimensional features when reducing high-dimensional data.\nWe begin by introducing a unifying theoretical framework for data reduction,\nconnecting PSPs with fundamental sampling principles from experimental design\nand Quasi-Monte Carlo. Through this framework, we then derive sparsity\nconditions under which the curse-of-dimensionality in data reduction can be\nlifted for our method. Next, we propose two algorithms for one-shot and\nsequential reduction via PSPs, both of which exploit big data subsampling and\nmajorization-minimization for efficient optimization. Finally, we demonstrate\nthe practical usefulness of PSPs in two real-world applications, the first for\ndata reduction in kernel learning, and the second for reducing Markov Chain\nMonte Carlo (MCMC) chains.\n",
"title": "Projected support points: a new method for high-dimensional data reduction"
}
| null | null |
[
"Statistics"
] | null | true | null |
13394
| null |
Validated
| null | null |
null |
{
"abstract": " We present AutonoVi:, a novel algorithm for autonomous vehicle navigation\nthat supports dynamic maneuvers and satisfies traffic constraints and norms.\nOur approach is based on optimization-based maneuver planning that supports\ndynamic lane-changes, swerving, and braking in all traffic scenarios and guides\nthe vehicle to its goal position. We take into account various traffic\nconstraints, including collision avoidance with other vehicles, pedestrians,\nand cyclists using control velocity obstacles. We use a data-driven approach to\nmodel the vehicle dynamics for control and collision avoidance. Furthermore,\nour trajectory computation algorithm takes into account traffic rules and\nbehaviors, such as stopping at intersections and stoplights, based on an\narc-spline representation. We have evaluated our algorithm in a simulated\nenvironment and tested its interactive performance in urban and highway driving\nscenarios with tens of vehicles, pedestrians, and cyclists. These scenarios\ninclude jaywalking pedestrians, sudden stops from high speeds, safely passing\ncyclists, a vehicle suddenly swerving into the roadway, and high-density\ntraffic where the vehicle must change lanes to progress more effectively.\n",
"title": "AutonoVi: Autonomous Vehicle Planning with Dynamic Maneuvers and Traffic Constraints"
}
| null | null | null | null | true | null |
13395
| null |
Default
| null | null |
null |
{
"abstract": " Stochastic Gradient Descent (SGD) is widely used in machine learning problems\nto efficiently perform empirical risk minimization, yet, in practice, SGD is\nknown to stall before reaching the actual minimizer of the empirical risk. SGD\nstalling has often been attributed to its sensitivity to the conditioning of\nthe problem; however, as we demonstrate, SGD will stall even when applied to a\nsimple linear regression problem with unity condition number for standard\nlearning rates. Thus, in this work, we numerically demonstrate and\nmathematically argue that stalling is a crippling and generic limitation of SGD\nand its variants in practice. Once we have established the problem of stalling,\nwe generalize an existing framework for hedging against its effects, which (1)\ndeters SGD and its variants from stalling, (2) still provides convergence\nguarantees, and (3) makes SGD and its variants more practical methods for\nminimization.\n",
"title": "On SGD's Failure in Practice: Characterizing and Overcoming Stalling"
}
| null | null | null | null | true | null |
13396
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we deal with composite rational functions having zeros and\npoles forming consecutive elements of an arithmetic progression. We also\ncorrect a result published earlier related to composite rational functions\nhaving a fixed number of zeros and poles.\n",
"title": "Composite Rational Functions and Arithmetic Progressions"
}
| null | null | null | null | true | null |
13397
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a geometry of interaction model for Mazza's multiport\ninteraction combinators, a graph-theoretic formalism which is able to\nfaithfully capture concurrent computation as embodied by process algebras like\nthe $\\pi$-calculus. The introduced model is based on token machines in which\nnot one but multiple tokens are allowed to traverse the underlying net at the\nsame time. We prove soundness and adequacy of the introduced model. The former\nis proved as a simulation result between the token machines one obtains along\nany reduction sequence. The latter is obtained by a fine analysis of\nconvergence, both in nets and in token machines.\n",
"title": "The Geometry of Concurrent Interaction: Handling Multiple Ports by Way of Multiple Tokens (Long Version)"
}
| null | null | null | null | true | null |
13398
| null |
Default
| null | null |
null |
{
"abstract": " We present the results of a systematic search for Lyman-alpha emitters (LAEs)\nat $6 \\lesssim z \\lesssim 7.6$ using the HST WFC3 Infrared Spectroscopic\nParallel (WISP) Survey. Our total volume over this redshift range is $\\sim 8\n\\times10^5$ Mpc$^3$, comparable to many of the narrowband surveys despite their\nlarger area coverage. We find two LAEs at $z=6.38$ and $6.44$ with line\nluminosities of L$_{\\mathrm{Ly}\\alpha} \\sim 4.7 \\times 10^{43}$ erg s$^{-1}$,\nputting them among the brightest LAEs discovered at these redshifts. Taking\nadvantage of the broad spectral coverage of WISP, we are able to rule out\nalmost all lower-redshift contaminants. The WISP LAEs have a high number\ndensity of $7.7\\times10^{-6}$ Mpc$^{-3}$. We argue that the LAEs reside in\nMpc-scale ionized bubbles that allow the Lyman-alpha photons to redshift out of\nresonance before encountering the neutral IGM. We discuss possible ionizing\nsources and conclude that the observed LAEs alone are not sufficient to ionize\nthe bubbles.\n",
"title": "A High Space Density of Luminous Lyman Alpha Emitters at z~6.5"
}
| null | null | null | null | true | null |
13399
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the geometry of the SYZ transform on a semi-flat\nLagrangian torus fibration. Our starting point is an investigation on the\nrelation between Lagrangian surgery of a pair of straight lines in a symplectic\n2-torus and extension of holomorphic vector bundles over the mirror elliptic\ncurve, via the SYZ transform for immersed Lagrangian multi-sections. This study\nleads us to a new notion of equivalence between objects in the immersed Fukaya\ncategory of a general compact symplectic manifold $(M, \\omega)$, under which\nthe immersed Floer cohomology is invariant; in particular, this provides an\nanswer to a question of Akaho-Joyce. Furthermore, if $M$ admits a Lagrangian\ntorus fibration over an integral affine manifold, we prove, under some\nadditional assumptions, that this new equivalence is mirror to isomorphism\nbetween holomorphic vector bundles over the dual torus fibration via the SYZ\ntransform.\n",
"title": "SYZ transforms for immersed Lagrangian multi-sections"
}
| null | null | null | null | true | null |
13400
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.