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": " Radio-loud high-redshift quasars (HRQs), although only a few of them are\nknown to date, are crucial for the studies of the growth of supermassive black\nholes (SMBHs) and the evolution of active galactic nuclei (AGN) at early\ncosmological epochs. Radio jets offer direct evidence of SMBHs, and their radio\nstructures can be studied with the highest angular resolution using Very Long\nBaseline Interferometry (VLBI). Here we report on the observations of three\nHRQs (J0131-0321, J0906+6930, J1026+2542) at z>5 using the Korean VLBI Network\nand VLBI Exploration of Radio Astrometry Arrays (together known as KaVA) with\nthe purpose of studying their pc-scale jet properties. The observations were\ncarried out at 22 and 43 GHz in 2016 January among the first-batch open-use\nexperiments of KaVA. The quasar J0906+6930 was detected at 22 GHz but not at 43\nGHz. The other two sources were not detected and upper limits to their compact\nradio emission are given. Archival VLBI imaging data and single-dish 15-GHz\nmonitoring light curve of J0906+6930 were also acquired as complementary\ninformation. J0906+6930 shows a moderate-level variability at 15 GHz. The radio\nimage is characterized by a core-jet structure with a total detectable size of\n~5 pc in projection. The brightness temperature, 1.9x10^{11} K, indicates\nrelativistic beaming of the jet. The radio properties of J0906+6930 are\nconsistent with a blazar. Follow-up VLBI observations will be helpful for\ndetermining its structural variation.\n",
"title": "J0906+6930: a radio-loud quasar in the early Universe"
}
| null | null | null | null | true | null |
6301
| null |
Default
| null | null |
null |
{
"abstract": " This work investigates the training of conditional random fields (CRFs) via\nthe stochastic dual coordinate ascent (SDCA) algorithm of Shalev-Shwartz and\nZhang (2016). SDCA enjoys a linear convergence rate and a strong empirical\nperformance for binary classification problems. However, it has never been used\nto train CRFs. Yet it benefits from an `exact' line search with a single\nmarginalization oracle call, unlike previous approaches. In this paper, we\nadapt SDCA to train CRFs, and we enhance it with an adaptive non-uniform\nsampling strategy based on block duality gaps. We perform experiments on four\nstandard sequence prediction tasks. SDCA demonstrates performances on par with\nthe state of the art, and improves over it on three of the four datasets, which\nhave in common the use of sparse features.\n",
"title": "Adaptive Stochastic Dual Coordinate Ascent for Conditional Random Fields"
}
| null | null | null | null | true | null |
6302
| null |
Default
| null | null |
null |
{
"abstract": " The availability of large idea repositories (e.g., the U.S. patent database)\ncould significantly accelerate innovation and discovery by providing people\nwith inspiration from solutions to analogous problems. However, finding useful\nanalogies in these large, messy, real-world repositories remains a persistent\nchallenge for either human or automated methods. Previous approaches include\ncostly hand-created databases that have high relational structure (e.g.,\npredicate calculus representations) but are very sparse. Simpler\nmachine-learning/information-retrieval similarity metrics can scale to large,\nnatural-language datasets, but struggle to account for structural similarity,\nwhich is central to analogy. In this paper we explore the viability and value\nof learning simpler structural representations, specifically, \"problem\nschemas\", which specify the purpose of a product and the mechanisms by which it\nachieves that purpose. Our approach combines crowdsourcing and recurrent neural\nnetworks to extract purpose and mechanism vector representations from product\ndescriptions. We demonstrate that these learned vectors allow us to find\nanalogies with higher precision and recall than traditional\ninformation-retrieval methods. In an ideation experiment, analogies retrieved\nby our models significantly increased people's likelihood of generating\ncreative ideas compared to analogies retrieved by traditional methods. Our\nresults suggest a promising approach to enabling computational analogy at scale\nis to learn and leverage weaker structural representations.\n",
"title": "Accelerating Innovation Through Analogy Mining"
}
| null | null | null | null | true | null |
6303
| null |
Default
| null | null |
null |
{
"abstract": " The object of this paper is to study $\\eta$-Ricci solitons on\n$(\\varepsilon)$-almost paracontact metric manifolds. We investigate\n$\\eta$-Ricci solitons in the case when its potential vector field is exactly\nthe characteristic vector field $\\xi$ of the $(\\varepsilon)$-almost paracontact\nmetric manifold and when the potential vector field is torse-forming. We also\nstudy Einstein-like and $(\\varepsilon)$-para Sasakian manifolds admitting\n$\\eta$-Ricci solitons. Finally we obtain some results for $\\eta$-Ricci solitons\non $(\\varepsilon)$-almost paracontact metric manifolds with a special view\ntowards parallel symmetric (0,2)-tensor fields.\n",
"title": "$η$-Ricci solitons in $(\\varepsilon)$-almost paracontact metric manifolds"
}
| null | null | null | null | true | null |
6304
| null |
Default
| null | null |
null |
{
"abstract": " We show how active transport of ions can be interpreted as an entropy\nfacilitated process. In this interpretation, the pore geometry through which\nsubstrates are transported can give rise to a driving force. This gives a\ndirect link between the geometry and the changes in Gibbs energy required.\nQuantifying the size of this effect for several proteins we find that the\nentropic contribution from the pore geometry is significant and we discuss how\nthe effect can be used to interpret variations in the affinity at the binding\nsite.\n",
"title": "Entropy facilitated active transport"
}
| null | null | null | null | true | null |
6305
| null |
Default
| null | null |
null |
{
"abstract": " In order to avoid well-know paradoxes associated with self-referential\ndefinitions, higher-order dependent type theories stratify the theory using a\ncountably infinite hierarchy of universes (also known as sorts), Type$_0$ :\nType$_1$ : $\\cdots$ . Such type systems are called cumulative if for any type\n$A$ we have that $A$ : Type$_{i}$ implies $A$ : Type$_{i+1}$. The predicative\ncalculus of inductive constructions (pCIC) which forms the basis of the Coq\nproof assistant, is one such system.\nIn this paper we present and establish the soundness of the predicative\ncalculus of cumulative inductive constructions (pCuIC) which extends the\ncumulativity relation to inductive types.\n",
"title": "Consistency of the Predicative Calculus of Cumulative Inductive Constructions (pCuIC)"
}
| null | null | null | null | true | null |
6306
| null |
Default
| null | null |
null |
{
"abstract": " A $(K, N, T, K_c)$ instance of the MDS-TPIR problem is comprised of $K$\nmessages and $N$ distributed servers. Each message is separately encoded\nthrough a $(K_c, N)$ MDS storage code. A user wishes to retrieve one message,\nas efficiently as possible, while revealing no information about the desired\nmessage index to any colluding set of up to $T$ servers. The fundamental limit\non the efficiency of retrieval, i.e., the capacity of MDS-TPIR is known only at\nthe extremes where either $T$ or $K_c$ belongs to $\\{1,N\\}$. The focus of this\nwork is a recent conjecture by Freij-Hollanti, Gnilke, Hollanti and Karpuk\nwhich offers a general capacity expression for MDS-TPIR. We prove that the\nconjecture is false by presenting as a counterexample a PIR scheme for the\nsetting $(K, N, T, K_c) = (2,4,2,2)$, which achieves the rate $3/5$, exceeding\nthe conjectured capacity, $4/7$. Insights from the counterexample lead us to\ncapacity characterizations for various instances of MDS-TPIR including all\ncases with $(K, N, T, K_c) = (2,N,T,N-1)$, where $N$ and $T$ can be arbitrary.\n",
"title": "Private Information Retrieval from MDS Coded Data with Colluding Servers: Settling a Conjecture by Freij-Hollanti et al."
}
| null | null |
[
"Computer Science"
] | null | true | null |
6307
| null |
Validated
| null | null |
null |
{
"abstract": " Deep Neural Networks (DNNs) and Convolutional Neural Networks (CNNs) are\nuseful for many practical tasks in machine learning. Synaptic weights, as well\nas neuron activation functions within the deep network are typically stored\nwith high-precision formats, e.g. 32 bit floating point. However, since storage\ncapacity is limited and each memory access consumes power, both storage\ncapacity and memory access are two crucial factors in these networks. Here we\npresent a method and present the ADaPTION toolbox to extend the popular deep\nlearning library Caffe to support training of deep CNNs with reduced numerical\nprecision of weights and activations using fixed point notation. ADaPTION\nincludes tools to measure the dynamic range of weights and activations. Using\nthe ADaPTION tools, we quantized several CNNs including VGG16 down to 16-bit\nweights and activations with only 0.8% drop in Top-1 accuracy. The\nquantization, especially of the activations, leads to increase of up to 50% of\nsparsity especially in early and intermediate layers, which we exploit to skip\nmultiplications with zero, thus performing faster and computationally cheaper\ninference.\n",
"title": "ADaPTION: Toolbox and Benchmark for Training Convolutional Neural Networks with Reduced Numerical Precision Weights and Activation"
}
| null | null | null | null | true | null |
6308
| null |
Default
| null | null |
null |
{
"abstract": " We present graph attention networks (GATs), novel neural network\narchitectures that operate on graph-structured data, leveraging masked\nself-attentional layers to address the shortcomings of prior methods based on\ngraph convolutions or their approximations. By stacking layers in which nodes\nare able to attend over their neighborhoods' features, we enable (implicitly)\nspecifying different weights to different nodes in a neighborhood, without\nrequiring any kind of costly matrix operation (such as inversion) or depending\non knowing the graph structure upfront. In this way, we address several key\nchallenges of spectral-based graph neural networks simultaneously, and make our\nmodel readily applicable to inductive as well as transductive problems. Our GAT\nmodels have achieved or matched state-of-the-art results across four\nestablished transductive and inductive graph benchmarks: the Cora, Citeseer and\nPubmed citation network datasets, as well as a protein-protein interaction\ndataset (wherein test graphs remain unseen during training).\n",
"title": "Graph Attention Networks"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6309
| null |
Validated
| null | null |
null |
{
"abstract": " The prevalence of online media has attracted researchers from various domains\nto explore human behavior and make interesting predictions. In this research,\nwe leverage heterogeneous social media data collected from various online\nplatforms to predict Taiwan's 2016 presidential election. In contrast to most\nexisting research, we take a \"signal\" view of heterogeneous information and\nadopt the Kalman filter to fuse multiple signals into daily vote predictions\nfor the candidates. We also consider events that influenced the election in a\nquantitative manner based on the so-called event study model that originated in\nthe field of financial research. We obtained the following interesting\nfindings. First, public opinions in online media dominate traditional polls in\nTaiwan election prediction in terms of both predictive power and timeliness.\nBut offline polls can still function on alleviating the sample bias of online\nopinions. Second, although online signals converge as election day approaches,\nthe simple Facebook \"Like\" is consistently the strongest indicator of the\nelection result. Third, most influential events have a strong connection to\ncross-strait relations, and the Chou Tzu-yu flag incident followed by the\napology video one day before the election increased the vote share of Tsai\nIng-Wen by 3.66%. This research justifies the predictive power of online media\nin politics and the advantages of information fusion. The combined use of the\nKalman filter and the event study method contributes to the data-driven\npolitical analytics paradigm for both prediction and attribution purposes.\n",
"title": "Social Media Would Not Lie: Prediction of the 2016 Taiwan Election via Online Heterogeneous Data"
}
| null | null | null | null | true | null |
6310
| null |
Default
| null | null |
null |
{
"abstract": " Technological advancements in the field of mobile devices and wearable\nsensors have helped overcome obstacles in the delivery of care, making it\npossible to deliver behavioral treatments anytime and anywhere. Increasingly\nthe delivery of these treatments is triggered by predictions of risk or\nengagement which may have been impacted by prior treatments. Furthermore the\ntreatments are often designed to have an impact on individuals over a span of\ntime during which subsequent treatments may be provided.\nHere we discuss our work on the design of a mobile health smoking cessation\nexperimental study in which two challenges arose. First the randomizations to\ntreatment should occur at times of stress and second the outcome of interest\naccrues over a period that may include subsequent treatment. To address these\nchallenges we develop the \"stratified micro-randomized trial,\" in which each\nindividual is randomized among treatments at times determined by predictions\nconstructed from outcomes to prior treatment and with randomization\nprobabilities depending on these outcomes. We define both conditional and\nmarginal proximal treatment effects. Depending on the scientific goal these\neffects may be defined over a period of time during which subsequent treatments\nmay be provided. We develop a primary analysis method and associated sample\nsize formulae for testing these effects.\n",
"title": "The stratified micro-randomized trial design: sample size considerations for testing nested causal effects of time-varying treatments"
}
| null | null |
[
"Statistics"
] | null | true | null |
6311
| null |
Validated
| null | null |
null |
{
"abstract": " The singular values of products of standard complex Gaussian random matrices,\nor sub-blocks of Haar distributed unitary matrices, have the property that\ntheir probability distribution has an explicit, structured form referred to as\na polynomial ensemble. It is furthermore the case that the corresponding\nbi-orthogonal system can be determined in terms of Meijer G-functions, and the\ncorrelation kernel given as an explicit double contour integral. It has\nrecently been shown that the Hermitised product $X_M \\cdots X_2 X_1A X_1^T\nX_2^T \\cdots X_M^T$, where each $X_i$ is a standard real complex Gaussian\nmatrix, and $A$ is real anti-symmetric shares exhibits analogous properties.\nHere we use the theory of spherical functions and transforms to present a\ntheory which, for even dimensions, includes these properties of the latter\nproduct as a special case. As an example we show that the theory also allows\nfor a treatment of this class of Hermitised product when the $X_i$ are chosen\nas sub-blocks of Haar distributed real orthogonal matrices.\n",
"title": "Multiplicative Convolution of Real Asymmetric and Real Antisymmetric Matrices"
}
| null | null | null | null | true | null |
6312
| null |
Default
| null | null |
null |
{
"abstract": " The Griffiths conjecture asserts that every ample vector bundle $E$ over a\ncompact complex manifold $S$ admits a hermitian metric with positive curvature\nin the sense of Griffiths. In this article we give a sufficient condition for a\npositive hermitian metric on $\\mathcal{O}_{\\mathbb{P}(E^*)}(1)$ to induce a\nGriffiths positive $L^2$-metric on the vector bundle $E$. This result suggests\nto study the relative Kähler-Ricci flow on $\\mathcal{O}_{\\mathbb{P}(E^*)}(1)$\nfor the fibration $\\mathbb{P}(E^*)\\to S$. We define a flow and give arguments\nfor the convergence.\n",
"title": "An approach to Griffiths conjecture"
}
| null | null | null | null | true | null |
6313
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning and deep learning in particular has advanced tremendously on\nperceptual tasks in recent years. However, it remains vulnerable against\nadversarial perturbations of the input that have been crafted specifically to\nfool the system while being quasi-imperceptible to a human. In this work, we\npropose to augment deep neural networks with a small \"detector\" subnetwork\nwhich is trained on the binary classification task of distinguishing genuine\ndata from data containing adversarial perturbations. Our method is orthogonal\nto prior work on addressing adversarial perturbations, which has mostly focused\non making the classification network itself more robust. We show empirically\nthat adversarial perturbations can be detected surprisingly well even though\nthey are quasi-imperceptible to humans. Moreover, while the detectors have been\ntrained to detect only a specific adversary, they generalize to similar and\nweaker adversaries. In addition, we propose an adversarial attack that fools\nboth the classifier and the detector and a novel training procedure for the\ndetector that counteracts this attack.\n",
"title": "On Detecting Adversarial Perturbations"
}
| null | null | null | null | true | null |
6314
| null |
Default
| null | null |
null |
{
"abstract": " In this position paper, we question the current practice of calculating\nevaluation metrics for recommender systems as single numbers (e.g. precision\np=.28 or mean absolute error MAE = 1.21). We argue that single numbers express\nonly average effectiveness over a usually rather long period (e.g. a year or\neven longer), which provides only a vague and static view of the data. We\npropose that recommender-system researchers should instead calculate metrics\nfor time-series such as weeks or months, and plot the results in e.g. a line\nchart. This way, results show how algorithms' effectiveness develops over time,\nand hence the results allow drawing more meaningful conclusions about how an\nalgorithm will perform in the future. In this paper, we explain our reasoning,\nprovide an example to illustrate our reasoning and present suggestions for what\nthe community should do next.\n",
"title": "It's Time to Consider \"Time\" when Evaluating Recommender-System Algorithms [Proposal]"
}
| null | null | null | null | true | null |
6315
| null |
Default
| null | null |
null |
{
"abstract": " The observation of metallic ground states in a variety of two-dimensional\nelectronic systems poses a fundamental challenge for the theory of electron\nfluids. Here, we analyze evidence for the existence of a regime, which we call\nthe \"anomalous metal regime,\" in diverse 2D superconducting systems driven\nthrough a quantum superconductor to metal transition (QSMT) by tuning physical\nparameters such as the magnetic field, the gate voltage in the case of systems\nwith a MOSFET geometry, or the degree of disorder. The principal\nphenomenological observation is that in the anomalous metal, as a function of\ndecreasing temperature, the resistivity first drops as if the system were\napproaching a superconducting ground state, but then saturates at low\ntemperatures to a value that can be orders of magnitude smaller than the Drude\nvalue. The anomalous metal also shows a giant positive magneto-resistance.\nThus, it behaves as if it were a \"failed superconductor.\" This behavior is\nobserved in a broad range of parameters. We moreover exhibit, by theoretical\nsolution of a model of superconducting grains embedded in a metallic matrix,\nthat as a matter of principle such anomalous metallic behavior can occur in the\nneighborhood of a QSMT. However, we also argue that the robustness and\nubiquitous nature of the observed phenomena are difficult to reconcile with any\nexisting theoretical treatment, and speculate about the character of a more\nfundamental theoretical framework.\n",
"title": "Anomalous metals -- failed superconductors"
}
| null | null | null | null | true | null |
6316
| null |
Default
| null | null |
null |
{
"abstract": " Autonomous Surface Vehicles (ASVs) provide an effective way to actualize\napplications such as environment monitoring, search and rescue, and scientific\nresearches. However, the conventional ASVs depends overly on the stored energy.\nHybrid Sailboat, mainly powered by the wind, can solve this problem by using an\nauxiliary propulsion system. The electric energy cost of Hybrid Sailboat needs\nto be optimized to achieve the ocean automatic cruise mission. Based on\nadjusted setting on sails and rudders, this paper seeks the optimal trajectory\nfor autonomic cruising to reduce the energy cost by changing the heading angle\nof sailing upwind. The experiment results validate the heading angle accounts\nfor energy cost and the trajectory with the best heading angle saves up to\n23.7% than other conditions. Furthermore, the energy-time line can be used to\npredict the energy cost for long-time sailing.\n",
"title": "Energy Optimization of Automatic Hybrid Sailboat"
}
| null | null | null | null | true | null |
6317
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we present a technique for using the bootstrap to estimate the\noperating characteristics and their variability for certain types of ensemble\nmethods. Bootstrapping a model can require a huge amount of work if the\ntraining data set is large. Fortunately in many cases the technique lets us\ndetermine the effect of infinite resampling without actually refitting a single\nmodel. We apply the technique to the study of meta-parameter selection for\nrandom forests. We demonstrate that alternatives to bootstrap aggregation and\nto considering \\sqrt{d} features to split each node, where d is the number of\nfeatures, can produce improvements in predictive accuracy.\n",
"title": "Estimating the Operating Characteristics of Ensemble Methods"
}
| null | null | null | null | true | null |
6318
| null |
Default
| null | null |
null |
{
"abstract": " Kinetic-range turbulence in magnetized plasmas and, in particular, in the\ncontext of solar-wind turbulence has been extensively investigated over the\npast decades via numerical simulations. Among others, one of the widely adopted\nreduced plasma model is the so-called hybrid-kinetic model, where the ions are\nfully kinetic and the electrons are treated as a neutralizing (inertial or\nmassless) fluid. Within the same model, different numerical methods and/or\napproaches to turbulence development have been employed. In the present work,\nwe present a comparison between two-dimensional hybrid-kinetic simulations of\nplasma turbulence obtained with two complementary approaches spanning about two\ndecades in wavenumber - from MHD inertial range to scales well below the ion\ngyroradius - with a state-of-the-art accuracy. One approach employs hybrid\nparticle-in-cell (HPIC) simulations of freely-decaying Alfvénic turbulence,\nwhereas the other consists of Eulerian hybrid Vlasov-Maxwell (HVM) simulations\nof turbulence continuously driven with partially-compressible large-scale\nfluctuations. Despite the completely different initialization and\ninjection/drive at large scales, the same properties of turbulent fluctuations\nat $k_\\perp\\rho_i\\gtrsim1$ are observed. The system indeed self-consistently\n\"reprocesses\" the turbulent fluctuations while they are cascading towards\nsmaller and smaller scales, in a way which actually depends on the plasma beta\nparameter. Small-scale turbulence has been found to be mainly populated by\nkinetic Alfvén wave (KAW) fluctuations for $\\beta\\geq1$, whereas KAW\nfluctuations are only sub-dominant for low-$\\beta$.\n",
"title": "Plasma turbulence at ion scales: a comparison between PIC and Eulerian hybrid-kinetic approaches"
}
| null | null | null | null | true | null |
6319
| null |
Default
| null | null |
null |
{
"abstract": " A lattice (d, k)-polytope is the convex hull of a set of points in dimension\nd whose coordinates are integers between 0 and k. Let {\\delta}(d, k) be the\nlargest diameter over all lattice (d, k)-polytopes. We develop a computational\nframework to determine {\\delta}(d, k) for small instances. We show that\n{\\delta}(3, 4) = 7 and {\\delta}(3, 5) = 9; that is, we verify for (d, k) = (3,\n4) and (3, 5) the conjecture whereby {\\delta}(d, k) is at most (k + 1)d/2 and\nis achieved, up to translation, by a Minkowski sum of lattice vectors.\n",
"title": "Computational determination of the largest lattice polytope diameter"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6320
| null |
Validated
| null | null |
null |
{
"abstract": " We report on an ion-optical system that serves as a microscope for ultracold\nground state and Rydberg atoms. The system is designed to achieve a\nmagnification of up to 1000 and a spatial resolution in the 100 nm range,\nthereby surpassing many standard imaging techniques for cold atoms. The\nmicroscope consists of four electrostatic lenses and a microchannel plate in\nconjunction with a delay line detector in order to achieve single particle\nsensitivity with high temporal and spatial resolution. We describe the design\nprocess of the microscope including ion-optical simulations of the imaging\nsystem and characterize aberrations and the resolution limit. Furthermore, we\npresent the experimental realization of the microscope in a cold atom setup and\ninvestigate its performance by patterned ionization with a structure size down\nto 2.7 {\\mu}m. The microscope meets the requirements for studying various\nmany-body effects, ranging from correlations in cold quantum gases up to\nRydberg molecule formation.\n",
"title": "A high resolution ion microscope for cold atoms"
}
| null | null | null | null | true | null |
6321
| null |
Default
| null | null |
null |
{
"abstract": " Dependency parsing is an important NLP task. A popular approach for\ndependency parsing is structured perceptron. Still, graph-based dependency\nparsing has the time complexity of $O(n^3)$, and it suffers from slow training.\nTo deal with this problem, we propose a parallel algorithm called parallel\nperceptron. The parallel algorithm can make full use of a multi-core computer\nwhich saves a lot of training time. Based on experiments we observe that\ndependency parsing with parallel perceptron can achieve 8-fold faster training\nspeed than traditional structured perceptron methods when using 10 threads, and\nwith no loss at all in accuracy.\n",
"title": "Lock-Free Parallel Perceptron for Graph-based Dependency Parsing"
}
| null | null | null | null | true | null |
6322
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\frak {F}$ be a class of group. A subgroup $A$ of a finite group $G$ is\nsaid to be $K$-$\\mathfrak{F}$-subnormal in $G$ if there is a subgroup chain\n$$A=A_{0} \\leq A_{1} \\leq \\cdots \\leq A_{n}=G$$ such that either $A_{i-1}\n\\trianglelefteq A_{i}$ or $A_{i}/(A_{i-1})_{A_{i}} \\in \\mathfrak{F}$ for all\n$i=1, \\ldots , n$. A formation $\\frak {F}$ is said to be $K$-lattice provided\nin every finite group $G$ the set of all its $K$-$\\mathfrak{F}$-subnormal\nsubgroups forms a sublattice of the lattice of all subgroups of $G$.\nIn this paper we consider some new applications of the theory of $K$-lattice\nformations. In particular, we prove the following\nTheorem A. Let $\\mathfrak{F}$ be a hereditary $K$-lattice saturated formation\ncontaining all nilpotent groups.\n(i) If every $\\mathfrak{F}$-critical subgroup $H$ of $G$ is\n$K$-$\\mathfrak{F}$-subnormal in $G$ with $H/F(H)\\in {\\mathfrak{F}}$, then\n$G/F(G)\\in {\\mathfrak{F}}$.\n(ii) If every Schmidt subgroup of $G$ is $K$-$\\mathfrak{F}$-subnormal in $G$,\nthen $G/G_{\\mathfrak{F}}$ is abelian.\n",
"title": "Finite groups with systems of $K$-$\\frak{F}$-subnormal subgroups"
}
| null | null | null | null | true | null |
6323
| null |
Default
| null | null |
null |
{
"abstract": " Assessing the impact of the individual actions performed by soccer players\nduring games is a crucial aspect of the player recruitment process.\nUnfortunately, most traditional metrics fall short in addressing this task as\nthey either focus on rare events like shots and goals alone or fail to account\nfor the context in which the actions occurred. This paper introduces a novel\nadvanced soccer metric for valuing any type of individual player action on the\npitch, be it with or without the ball. Our metric values each player action\nbased on its impact on the game outcome while accounting for the circumstances\nunder which the action happened. When applied to on-the-ball actions like\npasses, dribbles, and shots alone, our metric identifies Argentine forward\nLionel Messi, French teenage star Kylian Mbappé, and Belgian winger Eden\nHazard as the most effective players during the 2016/2017 season.\n",
"title": "Actions Speak Louder Than Goals: Valuing Player Actions in Soccer"
}
| null | null | null | null | true | null |
6324
| null |
Default
| null | null |
null |
{
"abstract": " M87, the active galaxy at the center of the Virgo cluster, is ideal for\nstudying the interaction of a supermassive black hole (SMBH) with a hot,\ngas-rich environment. A deep Chandra observation of M87 exhibits an\napproximately circular shock front (13 kpc radius, in projection) driven by the\nexpansion of the central cavity (filled by the SMBH with relativistic\nradio-emitting plasma) with projected radius $\\sim$1.9 kpc. We combine\nconstraints from X-ray and radio observations of M87 with a shock model to\nderive the properties of the outburst that created the 13 kpc shock. Principal\nconstraints for the model are 1) the measured Mach number ($M$$\\sim$1.2), 2)\nthe radius of the 13 kpc shock, and 3) the observed size of the central\ncavity/bubble (the radio-bright cocoon) that serves as the piston to drive the\nshock. We find an outburst of $\\sim$5$\\times$$10^{57}$ ergs that began about 12\nMyr ago and lasted $\\sim$2 Myr matches all the constraints. In this model,\n$\\sim$22% of the energy is carried by the shock as it expands. The remaining\n$\\sim$80% of the outburst energy is available to heat the core gas. More than\nhalf the total outburst energy initially goes into the enthalpy of the central\nbubble, the radio cocoon. As the buoyant bubble rises, much of its energy is\ntransferred to the ambient thermal gas. For an outburst repetition rate of\nabout 12 Myrs (the age of the outburst), 80% of the outburst energy is\nsufficient to balance the radiative cooling.\n",
"title": "Partitioning the Outburst Energy of a Low Eddington Accretion Rate AGN at the Center of an Elliptical Galaxy: the Recent 12 Myr History of the Supermassive Black Hole in M87"
}
| null | null |
[
"Physics"
] | null | true | null |
6325
| null |
Validated
| null | null |
null |
{
"abstract": " After decades of experimental, theoretical, and numerical research in fluid\ndynamics, many aspects of turbulence remain poorly understood. The main reason\nfor this is often attributed to the multiscale nature of turbulent flows, which\nposes a formidable challenge. There are, however, properties of these flows\nwhose roles and inter-connections have never been clarified fully. In this\narticle, we present a new connection between the pressure drop, viscous\ndissipation, and the turbulent energy spectrum, which, to the best of our\nknowledge, has never been established prior to our work. We use this finding to\nshow analytically that viscous dissipation in laminar pipe flows cannot\nincrease the temperature of the fluid, and to also reproduce qualitatively\nNikuradse's experimental results involving pressure drops in turbulent flows in\nrough pipes.\n",
"title": "On the missing link between pressure drop, viscous dissipation, and the turbulent energy spectrum"
}
| null | null | null | null | true | null |
6326
| null |
Default
| null | null |
null |
{
"abstract": " The local induction equation, or the binormal flow on space curves is a\nwell-known model of deformation of space curves as it describes the dynamics of\nvortex filaments, and the complex curvature is governed by the nonlinear\nSchrödinger equation. In this paper, we present its discrete analogue,\nnamely, a model of deformation of discrete space curves by the discrete\nnonlinear Schrödinger equation. We also present explicit formulas for both\nsmooth and discrete curves in terms of tau functions of the two-component KP\nhierarchy.\n",
"title": "Discrete Local Induction Equation"
}
| null | null | null | null | true | null |
6327
| null |
Default
| null | null |
null |
{
"abstract": " Let $T_{\\epsilon}$ be the lifespan for the solution to the Schrödinger\nequation on $\\mathbb{R}^d$ with a power nonlinearity $\\lambda |u|^{2\\theta/d}u$\n($\\lambda \\in \\mathbb{C}$, $0<\\theta<1$) and the initial data in the form\n$\\epsilon \\varphi(x)$. We provide a sharp lower bound estimate for\n$T_{\\epsilon}$ as $\\epsilon \\to +0$ which can be written explicitly by\n$\\lambda$, $d$, $\\theta$, $\\varphi$ and $\\epsilon$. This is an improvement of\nthe previous result by H.Sasaki [Adv. Diff. Eq. 14 (2009), 1021--1039].\n",
"title": "A sharp lower bound for the lifespan of small solutions to the Schrödinger equation with a subcritical power nonlinearity"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6328
| null |
Validated
| null | null |
null |
{
"abstract": " Classical CTL temporal logics are built over systems with interleaving model\nconcurrency. Many attempts are made to fight a state space explosion problem\n(for instance, compositional model checking). There are some methods of\nreduction of a state space based on independence of actions. However, in CSM\nmodel, which is based on coincidences rather than on interleaving, independence\nof actions cannot be defined. Therefore a state space reduction basing on\nidentical temporal consequences rather than on independence of action is\nproposed. The new reduction is not as good as for interleaving systems, because\nall successors of a state (in depth of two levels) must be obtained before a\nreduction may be applied. This leads to reduction of space required for\nrepresentation of a state space, but not in time of state space construction.\nYet much savings may occur in regular state spaces for CSM systems.\n",
"title": "State Space Reduction for Reachability Graph of CSM Automata"
}
| null | null | null | null | true | null |
6329
| null |
Default
| null | null |
null |
{
"abstract": " Information about the memory locations accessed by a program is, for\ninstance, required for program parallelisation and program verification.\nExisting inference techniques for this information provide only partial\nsolutions for the important class of array-manipulating programs. In this\npaper, we present a static analysis that infers the memory footprint of an\narray program in terms of permission pre- and postconditions as used, for\nexample, in separation logic. This formulation allows our analysis to handle\nconcurrent programs and produces specifications that can be used by\nverification tools. Our analysis expresses the permissions required by a loop\nvia maximum expressions over the individual loop iterations. These maximum\nexpressions are then solved by a novel maximum elimination algorithm, in the\nspirit of quantifier elimination. Our approach is sound and is implemented; an\nevaluation on existing benchmarks for memory safety of array programs\ndemonstrates accurate results, even for programs with complex access patterns\nand nested loops.\n",
"title": "Permission Inference for Array Programs"
}
| null | null | null | null | true | null |
6330
| null |
Default
| null | null |
null |
{
"abstract": " We address the problem of generating query suggestions to support users in\ncompleting their underlying tasks (which motivated them to search in the first\nplace). Given an initial query, these query suggestions should provide a\ncoverage of possible subtasks the user might be looking for. We propose a\nprobabilistic modeling framework that obtains keyphrases from multiple sources\nand generates query suggestions from these keyphrases. Using the test suites of\nthe TREC Tasks track, we evaluate and analyze each component of our model.\n",
"title": "Generating Query Suggestions to Support Task-Based Search"
}
| null | null | null | null | true | null |
6331
| null |
Default
| null | null |
null |
{
"abstract": " The Cosmic Axion Spin Precession Experiment (CASPEr) seeks to measure\noscillating torques on nuclear spins caused by axion or axion-like-particle\n(ALP) dark matter via nuclear magnetic resonance (NMR) techniques. A sample\nspin-polarized along a leading magnetic field experiences a resonance when the\nLarmor frequency matches the axion/ALP Compton frequency, generating precessing\ntransverse nuclear magnetization. Here we demonstrate a Spin-Exchange\nRelaxation-Free (SERF) magnetometer with sensitivity $\\approx 1~{\\rm\nfT/\\sqrt{Hz}}$ and an effective sensing volume of 0.1 $\\rm{cm^3}$ that may be\nuseful for NMR detection in CASPEr. A potential drawback of\nSERF-magnetometer-based NMR detection is the SERF's limited dynamic range. Use\nof a magnetic flux transformer to suppress the leading magnetic field is\nconsidered as a potential method to expand the SERF's dynamic range in order to\nprobe higher axion/ALP Compton frequencies.\n",
"title": "Application of Spin-Exchange Relaxation-Free Magnetometry to the Cosmic Axion Spin Precession Experiment"
}
| null | null |
[
"Physics"
] | null | true | null |
6332
| null |
Validated
| null | null |
null |
{
"abstract": " Quantum reactive scattering calculations are reported for the ultracold\nhydrogen-exchange reaction and its non-reactive atom-exchange isotopic\ncounterparts, proceeding from excited rotational states. It is shown that while\nthe geometric phase (GP) does not necessarily control the reaction to all final\nstates one can always find final states where it does. For the isotopic\ncounterpart reactions these states can be used to make a measurement of the GP\neffect by separately measuring the even and odd symmetry contributions, which\nexperimentally requires nuclear-spin final-state resolution. This follows from\nsymmetry considerations that make the even and odd identical-particle exchange\nsymmetry wavefunctions which include the GP locally equivalent to the opposite\nsymmetry wavefunctions which do not. This equivalence reflects the important\nrole discrete symmetries play in ultracold chemistry generally and highlights\nthe key role ultracold reactions can play in understanding fundamental aspects\nof chemical reactivity.\n",
"title": "Symmetry and the Geometric Phase in Ultracold Hydrogen-Exchange Reactions"
}
| null | null | null | null | true | null |
6333
| null |
Default
| null | null |
null |
{
"abstract": " The Subaru Strategic Program (SSP) is an ambitious multi-band survey using\nthe Hyper Suprime-Cam (HSC) on the Subaru telescope. The Wide layer of the SSP\nis both wide and deep, reaching a detection limit of i~26.0 mag. At these\ndepths, it is challenging to achieve accurate, unbiased, and consistent\nphotometry across all five bands. The HSC data are reduced using a pipeline\nthat builds on the prototype pipeline for the Large Synoptic Survey Telescope.\nWe have developed a Python-based, flexible framework to inject synthetic\ngalaxies into real HSC images called SynPipe. Here we explain the design and\nimplementation of SynPipe and generate a sample of synthetic galaxies to\nexamine the photometric performance of the HSC pipeline. For stars, we achieve\n1% photometric precision at i~19.0 mag and 6% precision at i~25.0 in the\ni-band. For synthetic galaxies with single-Sersic profiles, forced CModel\nphotometry achieves 13% photometric precision at i~20.0 mag and 18% precision\nat i~25.0 in the i-band. We show that both forced PSF and CModel photometry\nyield unbiased color estimates that are robust to seeing conditions. We\nidentify several caveats that apply to the version of HSC pipeline used for the\nfirst public HSC data release (DR1) that need to be taking into consideration.\nFirst, the degree to which an object is blended with other objects impacts the\noverall photometric performance. This is especially true for point sources.\nHighly blended objects tend to have larger photometric uncertainties,\nsystematically underestimated fluxes and slightly biased colors. Second, >20%\nof stars at 22.5< i < 25.0 mag can be misclassified as extended objects. Third,\nthe current CModel algorithm tends to strongly underestimate the half-light\nradius and ellipticity of galaxy with i>21.5 mag.\n",
"title": "Characterization and Photometric Performance of the Hyper Suprime-Cam Software Pipeline"
}
| null | null | null | null | true | null |
6334
| null |
Default
| null | null |
null |
{
"abstract": " We consider the estimation of hidden Markovian process by using information\ngeometry with respect to transition matrices. We consider the case when we use\nonly the histogram of $k$-memory data. Firstly, we focus on a partial\nobservation model with Markovian process and we show that the asymptotic\nestimation error of this model is given as the inverse of projective Fisher\ninformation of transition matrices. Next, we apply this result to the\nestimation of hidden Markovian process. We carefully discuss the equivalence\nproblem for hidden Markovian process on the tangent space. Then, we propose a\nnovel method to estimate hidden Markovian process.\n",
"title": "Information Geometry Approach to Parameter Estimation in Hidden Markov Models"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6335
| null |
Validated
| null | null |
null |
{
"abstract": " A nice differential-geometric framework for (non-abelian) higher gauge theory\nis provided by principal 2-bundles, i.e. categorified principal bundles. Their\ntotal spaces are Lie groupoids, local trivializations are kinds of Morita\nequivalences, and connections are Lie-2-algebra-valued 1-forms. In this\narticle, we construct explicitly the parallel transport of a connection on a\nprincipal 2-bundle. Parallel transport along a path is a Morita equivalence\nbetween the fibres over the end points, and parallel transport along a surface\nis an intertwiner between Morita equivalences. We prove that our constructions\nfit into the general axiomatic framework for categorified parallel transport\nand surface holonomy.\n",
"title": "Parallel transport in principal 2-bundles"
}
| null | null | null | null | true | null |
6336
| null |
Default
| null | null |
null |
{
"abstract": " In this report, we present a new reinforcement learning (RL) benchmark based\non the Sonic the Hedgehog (TM) video game franchise. This benchmark is intended\nto measure the performance of transfer learning and few-shot learning\nalgorithms in the RL domain. We also present and evaluate some baseline\nalgorithms on the new benchmark.\n",
"title": "Gotta Learn Fast: A New Benchmark for Generalization in RL"
}
| null | null | null | null | true | null |
6337
| null |
Default
| null | null |
null |
{
"abstract": " Observations of astrophysical objects such as galaxies are limited by various\nsources of random and systematic noise from the sky background, the optical\nsystem of the telescope and the detector used to record the data. Conventional\ndeconvolution techniques are limited in their ability to recover features in\nimaging data by the Shannon-Nyquist sampling theorem. Here we train a\ngenerative adversarial network (GAN) on a sample of $4,550$ images of nearby\ngalaxies at $0.01<z<0.02$ from the Sloan Digital Sky Survey and conduct\n$10\\times$ cross validation to evaluate the results. We present a method using\na GAN trained on galaxy images that can recover features from artificially\ndegraded images with worse seeing and higher noise than the original with a\nperformance which far exceeds simple deconvolution. The ability to better\nrecover detailed features such as galaxy morphology from low-signal-to-noise\nand low angular resolution imaging data significantly increases our ability to\nstudy existing data sets of astrophysical objects as well as future\nobservations with observatories such as the Large Synoptic Sky Telescope (LSST)\nand the Hubble and James Webb space telescopes.\n",
"title": "Generative Adversarial Networks recover features in astrophysical images of galaxies beyond the deconvolution limit"
}
| null | null | null | null | true | null |
6338
| null |
Default
| null | null |
null |
{
"abstract": " Flood risk changes in time and is influenced by both natural and\nsocio-economic trends and interactions. In Europe, previous studies of\nhistorical flood losses corrected for demographic and economic growth\n(\"normalized\") have been limited in temporal and spatial extent, leading to an\nincomplete representation in trends of losses over time. In this study we\nutilize a gridded reconstruction of flood exposure in 37 European countries and\na new database of damaging floods since 1870. Our results indicate that since\n1870 there has been an increase in annually inundated area and number of\npersons affected, contrasted by a substantial decrease in flood fatalities,\nafter correcting for change in flood exposure. For more recent decades we also\nfound a considerable decline in financial losses per year. We estimate,\nhowever, that there is large underreporting of smaller floods beyond most\nrecent years, and show that underreporting has a substantial impact on observed\ntrends.\n",
"title": "Trends in European flood risk over the past 150 years"
}
| null | null | null | null | true | null |
6339
| null |
Default
| null | null |
null |
{
"abstract": " The central dogma of molecular biology is the principal framework for\nunderstanding how nucleic acid information is propagated and used by living\nsystems to create complex biomolecules. Here, by integrating the structural and\ndynamic paradigms of DNA nanotechnology, we present a rationally designed\nsynthetic platform which functions in an analogous manner to create complex DNA\nnanostructures. Starting from one type of DNA nanostructure, DNA strand\ndisplacement circuits were designed to interact and pass along the information\nencoded in the initial structure to mediate the self-assembly of a different\ntype of structure, the final output structure depending on the type of circuit\ntriggered. Using this concept of a DNA structure \"trans-assembling\" a different\nDNA structure through non-local strand displacement circuitry, four different\nschemes were implemented. Specifically, 1D ladder and 2D double-crossover (DX)\nlattices were designed to kinetically trigger DNA circuits to activate\npolymerization of either ring structures or another type of DX lattice under\nenzyme-free, isothermal conditions. In each scheme, the desired multilayer\nreaction pathway was activated, among multiple possible pathways, ultimately\nleading to the downstream self-assembly of the correct output structure.\n",
"title": "Kinetic Trans-assembly of DNA Nanostructures"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
6340
| null |
Validated
| null | null |
null |
{
"abstract": " We generalize the bridge between analysis and synthesis estimators by Elad,\nMilanfar and Rubinstein (2007) to rank deficient cases. This is a starting\npoint for the study of the connection between analysis and synthesis for total\nvariation regularized estimators. In particular, the case of first order total\nvariation regularized estimators over general graphs and their synthesis form\nare studied.\nWe give a definition of the discrete graph derivative operator based on the\nnotion of line graph and provide examples of the synthesis form of\n$k^{\\text{th}}$ order total variation regularized estimators over a range of\ngraphs.\n",
"title": "Synthesis and analysis in total variation regularization"
}
| null | null | null | null | true | null |
6341
| null |
Default
| null | null |
null |
{
"abstract": " We compute the exact norms of the Leray transforms for a family\n$\\mathcal{S}_{\\beta}$ of unbounded hypersurfaces in two complex dimensions. The\n$\\mathcal{S}_{\\beta}$ generalize the Heisenberg group, and provide local\nprojective approximations to any smooth, strongly $\\mathbb{C}$-convex\nhypersurface $\\mathcal{S}_{\\beta}$ to two orders of tangency. This work is then\nexamined in the context of projective dual $CR$-structures and the\ncorresponding pair of canonical dual Hardy spaces associated to\n$\\mathcal{S}_{\\beta}$, leading to a universal description of the Leray\ntransform and a factorization of the transform through orthogonal projection\nonto the conjugate dual Hardy space.\n",
"title": "The Leray transform: factorization, dual $CR$ structures and model hypersurfaces in $\\mathbb{C}\\mathbb{P}^2$"
}
| null | null | null | null | true | null |
6342
| null |
Default
| null | null |
null |
{
"abstract": " This paper gives the definitions of an extra superincreasing sequence and an\nanomalous subset sum, and proposes a fast quantum-safe asymmetric cryptosystem\ncalled JUOAN2. The new cryptosystem is based on an additive multivariate\npermutation problem (AMPP) and an anomalous subset sum problem (ASSP) which\nparallel a multivariate polynomial problem and a shortest vector problem\nrespectively, and composed of a key generator, an encryption algorithm, and a\ndecryption algorithm. The authors analyze the security of the new cryptosystem\nagainst the Shamir minima accumulation point attack and the LLL lattice basis\nreduction attack, and prove it to be semantically secure (namely IND-CPA) on\nthe assumption that AMPP and ASSP have no subexponential time solutions.\nParticularly, the analysis shows that the new cryptosystem has the potential to\nbe resistant to quantum computing attack, and is especially suitable to the\nsecret communication between two mobile terminals in maneuvering field\noperations under any weather. At last, an example explaining the correctness of\nthe new cryptosystem is given.\n",
"title": "A Fast Quantum-safe Asymmetric Cryptosystem Using Extra Superincreasing Sequences"
}
| null | null | null | null | true | null |
6343
| null |
Default
| null | null |
null |
{
"abstract": " Knowledge transfer impacts the performance of deep learning -- the state of\nthe art for image classification tasks, including automated melanoma screening.\nDeep learning's greed for large amounts of training data poses a challenge for\nmedical tasks, which we can alleviate by recycling knowledge from models\ntrained on different tasks, in a scheme called transfer learning. Although much\nof the best art on automated melanoma screening employs some form of transfer\nlearning, a systematic evaluation was missing. Here we investigate the presence\nof transfer, from which task the transfer is sourced, and the application of\nfine tuning (i.e., retraining of the deep learning model after transfer). We\nalso test the impact of picking deeper (and more expensive) models. Our results\nfavor deeper models, pre-trained over ImageNet, with fine-tuning, reaching an\nAUC of 80.7% and 84.5% for the two skin-lesion datasets evaluated.\n",
"title": "Knowledge Transfer for Melanoma Screening with Deep Learning"
}
| null | null | null | null | true | null |
6344
| null |
Default
| null | null |
null |
{
"abstract": " For a primitive Dirichlet character $\\chi$ modulo $q$, we define\n$M(\\chi)=\\max_{t } |\\sum_{n \\leq t} \\chi(n)|$. In this paper, we study this\nquantity for characters of a fixed odd order $g\\geq 3$. Our main result\nprovides a further improvement of the classical Pólya-Vinogradov inequality\nin this case. More specifically, we show that for any such character $\\chi$ we\nhave $$M(\\chi)\\ll_{\\varepsilon} \\sqrt{q}(\\log q)^{1-\\delta_g}(\\log\\log\nq)^{-1/4+\\varepsilon},$$ where $\\delta_g := 1-\\frac{g}{\\pi}\\sin(\\pi/g)$. This\nimproves upon the works of Granville and Soundararajan and of Goldmakher.\nFurthermore, assuming the Generalized Riemann hypothesis (GRH) we prove that $$\nM(\\chi) \\ll \\sqrt{q} \\left(\\log_2 q\\right)^{1-\\delta_g} \\left(\\log_3\nq\\right)^{-\\frac{1}{4}}\\left(\\log_4 q\\right)^{O(1)}, $$ where $\\log_j$ is the\n$j$-th iterated logarithm. We also show unconditionally that this bound is best\npossible (up to a power of $\\log_4 q$). One of the key ingredients in the proof\nof the upper bounds is a new Halász-type inequality for logarithmic mean\nvalues of completely multiplicative functions, which might be of independent\ninterest.\n",
"title": "Large odd order character sums and improvements of the Pólya-Vinogradov inequality"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6345
| null |
Validated
| null | null |
null |
{
"abstract": " Motivated by geometric problems in signal processing, computer vision, and\nstructural biology, we study a class of orbit recovery problems where we\nobserve very noisy copies of an unknown signal, each acted upon by a random\nelement of some group (such as Z/p or SO(3)). The goal is to recover the orbit\nof the signal under the group action in the high-noise regime. This generalizes\nproblems of interest such as multi-reference alignment (MRA) and the\nreconstruction problem in cryo-electron microscopy (cryo-EM). We obtain\nmatching lower and upper bounds on the sample complexity of these problems in\nhigh generality, showing that the statistical difficulty is intricately\ndetermined by the invariant theory of the underlying symmetry group.\nIn particular, we determine that for cryo-EM with noise variance $\\sigma^2$\nand uniform viewing directions, the number of samples required scales as\n$\\sigma^6$. We match this bound with a novel algorithm for ab initio\nreconstruction in cryo-EM, based on invariant features of degree at most 3. We\nfurther discuss how to recover multiple molecular structures from heterogeneous\ncryo-EM samples.\n",
"title": "Estimation under group actions: recovering orbits from invariants"
}
| null | null | null | null | true | null |
6346
| null |
Default
| null | null |
null |
{
"abstract": " The pyrochlore magnet $\\rm Yb_2Ti_2O_7$ has been proposed as a quantum spin\nice candidate, a spin liquid state expected to display emergent quantum\nelectrodynamics with gauge photons among its elementary excitations. However,\n$\\rm Yb_2Ti_2O_7$'s ground state is known to be very sensitive to its precise\nstoichiometry. Powder samples, produced by solid state synthesis at relatively\nlow temperatures, tend to be stoichiometric, while single crystals grown from\nthe melt tend to display weak \"stuffing\" wherein $\\mathrm{\\sim 2\\%}$ of the\n$\\mathrm{Yb^{3+}}$, normally at the $A$ site of the $A_2B_2O_7$ pyrochlore\nstructure, reside as well at the $B$ site. In such samples $\\mathrm{Yb^{3+}}$\nions should exist in defective environments at low levels, and be subjected to\ncrystalline electric fields (CEFs) very different from those at the\nstoichiometric $A$ sites. New neutron scattering measurements of\n$\\mathrm{Yb^{3+}}$ in four compositions of $\\rm Yb_{2+x}Ti_{2-x}O_{7-y}$, show\nthe spectroscopic signatures for these defective $\\mathrm{Yb^{3+}}$ ions and\nexplicitly demonstrate that the spin anisotropy of the $\\mathrm{Yb^{3+}}$\nmoment changes from XY-like for stoichiometric $\\mathrm{Yb^{3+}}$, to\nIsing-like for \"stuffed\" $B$ site $\\mathrm{Yb^{3+}}$, or for $A$ site\n$\\mathrm{Yb^{3+}}$ in the presence of an oxygen vacancy.\n",
"title": "Crystal field excitations from $\\mathrm{Yb^{3+}}$ ions at defective sites in highly stuffed $\\rm Yb_2Ti_2O_7$"
}
| null | null | null | null | true | null |
6347
| null |
Default
| null | null |
null |
{
"abstract": " We present a neurosymbolic framework for the lifelong learning of algorithmic\ntasks that mix perception and procedural reasoning. Reusing high-level concepts\nacross domains and learning complex procedures are key challenges in lifelong\nlearning. We show that a program synthesis approach that combines gradient\ndescent with combinatorial search over programs can be a more effective\nresponse to these challenges than purely neural methods. Our framework, called\nHOUDINI, represents neural networks as strongly typed, differentiable\nfunctional programs that use symbolic higher-order combinators to compose a\nlibrary of neural functions. Our learning algorithm consists of: (1) a symbolic\nprogram synthesizer that performs a type-directed search over parameterized\nprograms, and decides on the library functions to reuse, and the architectures\nto combine them, while learning a sequence of tasks; and (2) a neural module\nthat trains these programs using stochastic gradient descent. We evaluate\nHOUDINI on three benchmarks that combine perception with the algorithmic tasks\nof counting, summing, and shortest-path computation. Our experiments show that\nHOUDINI transfers high-level concepts more effectively than traditional\ntransfer learning and progressive neural networks, and that the typed\nrepresentation of networks significantly accelerates the search.\n",
"title": "HOUDINI: Lifelong Learning as Program Synthesis"
}
| null | null | null | null | true | null |
6348
| null |
Default
| null | null |
null |
{
"abstract": " Though deep neural networks have achieved state-of-the-art performance in\nvisual classification, recent studies have shown that they are all vulnerable\nto the attack of adversarial examples. Small and often imperceptible\nperturbations to the input images are sufficient to fool the most powerful deep\nneural networks. Various defense methods have been proposed to address this\nissue. However, they either require knowledge on the process of generating\nadversarial examples, or are not robust against new attacks specifically\ndesigned to penetrate the existing defense. In this work, we introduce\nkey-based network, a new detection-based defense mechanism to distinguish\nadversarial examples from normal ones based on error correcting output codes,\nusing the binary code vectors produced by multiple binary classifiers applied\nto randomly chosen label-sets as signatures to match normal images and reject\nadversarial examples. In contrast to existing defense methods, the proposed\nmethod does not require knowledge of the process for generating adversarial\nexamples and can be applied to defend against different types of attacks. For\nthe practical black-box and gray-box scenarios, where the attacker does not\nknow the encoding scheme, we show empirically that key-based network can\neffectively detect adversarial examples generated by several state-of-the-art\nattacks.\n",
"title": "Detecting Adversarial Examples via Key-based Network"
}
| null | null | null | null | true | null |
6349
| null |
Default
| null | null |
null |
{
"abstract": " The secrecy of a distributed-storage system for passwords is studied. The\nencoder, Alice, observes a length-n password and describes it using two hints,\nwhich she stores in different locations. The legitimate receiver, Bob, observes\nboth hints. In one scenario the requirement is that the expected number of\nguesses it takes Bob to guess the password approach one as n tends to infinity,\nand in the other that the expected size of the shortest list that Bob must form\nto guarantee that it contain the password approach one. The eavesdropper, Eve,\nsees only one of the hints. Assuming that Alice cannot control which hints Eve\nobserves, the largest normalized (by n) exponent that can be guaranteed for the\nexpected number of guesses it takes Eve to guess the password is characterized\nfor each scenario. Key to the proof are new results on Arikan's guessing and\nBunte and Lapidoth's task-encoding problem; in particular, the paper\nestablishes a close relation between the two problems. A rate-distortion\nversion of the model is also discussed, as is a generalization that allows for\nAlice to produce {\\delta} (not necessarily two) hints, for Bob to observe {\\nu}\n(not necessarily two) of the hints, and for Eve to observe {\\eta} (not\nnecessarily one) of the hints. The generalized model is robust against {\\delta}\n- {\\nu} disk failures.\n",
"title": "Guessing Attacks on Distributed-Storage Systems"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
6350
| null |
Validated
| null | null |
null |
{
"abstract": " In this work, we calculate the convergence rate of the finite difference\napproximation for a class of nonlocal fracture models. We consider two point\nforce interactions characterized by a double well potential. We show the\nexistence of a evolving displacement field in Hölder space with Hölder\nexponent $\\gamma \\in (0,1]$. The rate of convergence of the finite difference\napproximation depends on the factor $C_s h^\\gamma/\\epsilon^2$ where $\\epsilon$\ngives the length scale of nonlocal interaction, $h$ is the discretization\nlength and $C_s$ is the maximum of Hölder norm of the solution and its second\nderivatives during the evolution. It is shown that the rate of convergence\nholds for both the forward Euler scheme as well as general single step implicit\nschemes. A stability result is established for the semi-discrete approximation.\nThe Hölder continuous evolutions are seen to converge to a brittle fracture\nevolution in the limit of vanishing nonlocality.\n",
"title": "Numerical analysis of nonlocal fracture models in Hölder space"
}
| null | null | null | null | true | null |
6351
| null |
Default
| null | null |
null |
{
"abstract": " The sparse matrix estimation problem consists of estimating the distribution\nof an $n\\times n$ matrix $Y$, from a sparsely observed single instance of this\nmatrix where the entries of $Y$ are independent random variables. This captures\na wide array of problems; special instances include matrix completion in the\ncontext of recommendation systems, graphon estimation, and community detection\nin (mixed membership) stochastic block models. Inspired by classical\ncollaborative filtering for recommendation systems, we propose a novel\niterative, collaborative filtering-style algorithm for matrix estimation in\nthis generic setting. We show that the mean squared error (MSE) of our\nestimator converges to $0$ at the rate of $O(d^2 (pn)^{-2/5})$ as long as\n$\\omega(d^5 n)$ random entries from a total of $n^2$ entries of $Y$ are\nobserved (uniformly sampled), $\\mathbb{E}[Y]$ has rank $d$, and the entries of\n$Y$ have bounded support. The maximum squared error across all entries\nconverges to $0$ with high probability as long as we observe a little more,\n$\\Omega(d^5 n \\ln^2(n))$ entries. Our results are the best known sample\ncomplexity results in this generality.\n",
"title": "Iterative Collaborative Filtering for Sparse Matrix Estimation"
}
| null | null | null | null | true | null |
6352
| null |
Default
| null | null |
null |
{
"abstract": " In this short paper, we formulate parameter estimation for finite mixture\nmodels in the context of discrete optimal transportation with convex\nregularization. The proposed framework unifies hard and soft clustering methods\nfor general mixture models. It also generalizes the celebrated\n$k$\\nobreakdash-means and expectation-maximization algorithms in relation to\nassociated Bregman divergences when applied to exponential family mixture\nmodels.\n",
"title": "Parameter Estimation in Finite Mixture Models by Regularized Optimal Transport: A Unified Framework for Hard and Soft Clustering"
}
| null | null | null | null | true | null |
6353
| null |
Default
| null | null |
null |
{
"abstract": " Compartmental equations are primary tools in disease spreading studies. Their\npredictions are accurate for large populations but disagree with empirical and\nsimulated data for finite populations, where uncertainties become a relevant\nfactor. Starting from the agent-based approach, we investigate the role of\nuncertainties and autocorrelation functions in SIS epidemic model, including\ntheir relationship with epidemiological variables. We find new differential\nequations that take uncertainties into account. The findings provide improved\npredictions to the SIS model and it can offer new insights for emerging\ndiseases.\n",
"title": "Robust parameter determination in epidemic models with analytical descriptions of uncertainties"
}
| null | null | null | null | true | null |
6354
| null |
Default
| null | null |
null |
{
"abstract": " The surface energy of a magnetic Domain Wall (DW) strongly affects its static\nand dynamic behaviours. However, this effect was seldom directly observed and\nmany related phenomena have not been well understood. Moreover, a reliable\nmethod to quantify the DW surface energy is still missing. Here, we report a\nseries of experiments in which the DW surface energy becomes a dominant\nparameter. We observed that a semicircular magnetic domain bubble could\nspontaneously collapse under the Laplace pressure induced by DW surface energy.\nWe further demonstrated that the surface energy could lead to a geometrically\ninduced pinning when the DW propagates in a Hall cross or from a nanowire into\na nucleation pad. Based on these observations, we developed two methods to\nquantify the DW surface energy, which could be very helpful to estimate\nintrinsic parameters such as Dzyaloshinskii-Moriya Interactions (DMI) or\nexchange stiffness in magnetic ultra-thin films.\n",
"title": "Direct observation of domain wall surface tension by deflating or inflating a magnetic bubble"
}
| null | null |
[
"Physics"
] | null | true | null |
6355
| null |
Validated
| null | null |
null |
{
"abstract": " Using the Fenchel-Eggleston theorem for convex hulls (an extension of the\nCaratheodory theorem), we prove that any likelihood can be maximized by either\na dark matter 1- speed distribution $F(v)$ in Earth's frame or 2- Galactic\nvelocity distribution $f^{\\rm gal}(\\vec{u})$, consisting of a sum of delta\nfunctions. The former case applies only to time-averaged rate measurements and\nthe maximum number of delta functions is $({\\mathcal N}-1)$, where ${\\mathcal\nN}$ is the total number of data entries. The second case applies to any\nharmonic expansion coefficient of the time-dependent rate and the maximum\nnumber of terms is ${\\mathcal N}$. Using time-averaged rates, the\naforementioned form of $F(v)$ results in a piecewise constant unmodulated halo\nfunction $\\tilde\\eta^0_{BF}(v_{\\rm min})$ (which is an integral of the speed\ndistribution) with at most $({\\mathcal N}-1)$ downward steps. The authors had\npreviously proven this result for likelihoods comprised of at least one\nextended likelihood, and found the best-fit halo function to be unique. This\nuniqueness, however, cannot be guaranteed in the more general analysis applied\nto arbitrary likelihoods. Thus we introduce a method for determining whether\nthere exists a unique best-fit halo function, and provide a procedure for\nconstructing either a pointwise confidence band, if the best-fit halo function\nis unique, or a degeneracy band, if it is not. Using measurements of modulation\namplitudes, the aforementioned form of $f^{\\rm gal}(\\vec{u})$, which is a sum\nof Galactic streams, yields a periodic time-dependent halo function\n$\\tilde\\eta_{BF}(v_{\\rm min}, t)$ which at any fixed time is a piecewise\nconstant function of $v_{\\rm min}$ with at most ${\\mathcal N}$ downward steps.\nIn this case, we explain how to construct pointwise confidence and degeneracy\nbands from the time-averaged halo function. Finally, we show that requiring an\nisotropic ...\n",
"title": "Unified Halo-Independent Formalism From Convex Hulls for Direct Dark Matter Searches"
}
| null | null |
[
"Physics"
] | null | true | null |
6356
| null |
Validated
| null | null |
null |
{
"abstract": " We study finite-size fluctuations in a network of spiking deterministic\nneurons coupled with non-uniform synaptic coupling. We generalize a previously\ndeveloped theory of finite size effects for uniform globally coupled neurons.\nIn the uniform case, mean field theory is well defined by averaging over the\nnetwork as the number of neurons in the network goes to infinity. However, for\nnonuniform coupling it is no longer possible to average over the entire network\nif we are interested in fluctuations at a particular location within the\nnetwork. We show that if the coupling function approaches a continuous function\nin the infinite system size limit then an average over a local neighborhood can\nbe defined such that mean field theory is well defined for a spatially\ndependent field. We then derive a perturbation expansion in the inverse system\nsize around the mean field limit for the covariance of the input to a neuron\n(synaptic drive) and firing rate fluctuations due to dynamical deterministic\nfinite-size effects.\n",
"title": "Finite size effects for spiking neural networks with spatially dependent coupling"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
6357
| null |
Validated
| null | null |
null |
{
"abstract": " We study static distributions of ferrofluid submitted to non-uniform magnetic\nfields. We show how the normal-field instability is modified in the presence of\na weak magnetic field gradient. Then we consider a ferrofluid droplet and show\nhow the gradient affects its shape. A rich phase transitions phenomenology is\nfound. We also investigate the creation of droplets by successive splits when a\nmagnet is vertically approached from below and derive theoretical expressions\nwhich are solved numerically to obtain the number of droplets and their aspect\nratio as function of the field configuration. A quantitative comparison is\nperformed with previous experimental results, as well as with our own\nexperiments, and yields good agreement with the theoretical modeling.\n",
"title": "Shape and fission instabilities of ferrofluids in non-uniform magnetic fields"
}
| null | null | null | null | true | null |
6358
| null |
Default
| null | null |
null |
{
"abstract": " Information that is stored in an encrypted format is, by definition, usually\nnot amenable to statistical analysis or machine learning methods. In this paper\nwe present detailed analysis of coordinate and accelerated gradient descent\nalgorithms which are capable of fitting least squares and penalised ridge\nregression models, using data encrypted under a fully homomorphic encryption\nscheme. Gradient descent is shown to dominate in terms of encrypted\ncomputational speed, and theoretical results are proven to give parameter\nbounds which ensure correctness of decryption. The characteristics of encrypted\ncomputation are empirically shown to favour a non-standard acceleration\ntechnique. This demonstrates the possibility of approximating conventional\nstatistical regression methods using encrypted data without compromising\nprivacy.\n",
"title": "Encrypted accelerated least squares regression"
}
| null | null | null | null | true | null |
6359
| null |
Default
| null | null |
null |
{
"abstract": " The large hierarchy between the Planck scale and the weak scale can be\nexplained by the dynamical breaking of supersymmetry in strongly coupled gauge\ntheories. Similarly, the hierarchy between the Planck scale and the energy\nscale of inflation may also originate from strong dynamics, which dynamically\ngenerate the inflaton potential. We present a model of the hidden sector which\nunifies these two ideas, i.e., in which the scales of inflation and\nsupersymmetry breaking are provided by the dynamics of the same gauge group.\nThe resultant inflation model is chaotic inflation with a fractional power-law\npotential in accord with the upper bound on the tensor-to-scalar ratio. The\nsupersymmetry breaking scale can be much smaller than the inflation scale, so\nthat the solution to the large hierarchy problem of the weak scale remains\nintact. As an intrinsic feature of our model, we find that the sgoldstino,\nwhich might disturb the inflationary dynamics, is automatically stabilized\nduring inflation by dynamically generated corrections in the strongly coupled\nsector. This renders our model a field-theoretical realization of what is\nsometimes referred to as sgoldstino-less inflation.\n",
"title": "Unified Model of Chaotic Inflation and Dynamical Supersymmetry Breaking"
}
| null | null | null | null | true | null |
6360
| null |
Default
| null | null |
null |
{
"abstract": " In many scenarios of a language identification task, the user will specify a\nsmall set of languages which he/she can speak instead of a large set of all\npossible languages. We want to model such prior knowledge into the way we train\nour neural networks, by replacing the commonly used softmax loss function with\na novel loss function named tuplemax loss. As a matter of fact, a typical\nlanguage identification system launched in North America has about 95% users\nwho could speak no more than two languages. Using the tuplemax loss, our system\nachieved a 2.33% error rate, which is a relative 39.4% improvement over the\n3.85% error rate of standard softmax loss method.\n",
"title": "Tuplemax Loss for Language Identification"
}
| null | null | null | null | true | null |
6361
| null |
Default
| null | null |
null |
{
"abstract": " Example-based mesh deformation methods are powerful tools for realistic shape\nediting. However, existing techniques typically combine all the example\ndeformation modes, which can lead to overfitting, i.e. using a overly\ncomplicated model to explain the user-specified deformation. This leads to\nimplausible or unstable deformation results, including unexpected global\nchanges outside the region of interest. To address this fundamental limitation,\nwe propose a sparse blending method that automatically selects a smaller number\nof deformation modes to compactly describe the desired deformation. This along\nwith a suitably chosen deformation basis including spatially localized\ndeformation modes leads to significant advantages, including more meaningful,\nreliable, and efficient deformations because fewer and localized deformation\nmodes are applied. To cope with large rotations, we develop a simple but\neffective representation based on polar decomposition of deformation gradients,\nwhich resolves the ambiguity of large global rotations using an\nas-consistent-as-possible global optimization. This simple representation has a\nclosed form solution for derivatives, making it efficient for sparse localized\nrepresentation and thus ensuring interactive performance. Experimental results\nshow that our method outperforms state-of-the-art data-driven mesh deformation\nmethods, for both quality of results and efficiency.\n",
"title": "Sparse Data Driven Mesh Deformation"
}
| null | null | null | null | true | null |
6362
| null |
Default
| null | null |
null |
{
"abstract": " We study the computational complexity of short sentences in Presburger\narithmetic (Short-PA). Here by \"short\" we mean sentences with a bounded number\nof variables, quantifiers, inequalities and Boolean operations; the input\nconsists only of the integer coefficients involved in the linear inequalities.\nWe prove that satisfiability of Short-PA sentences with $m+2$ alternating\nquantifiers is $\\Sigma_{P}^m$-complete or $\\Pi_{P}^m$-complete, when the first\nquantifier is $\\exists$ or $\\forall$, respectively. Counting versions and\nrestricted systems are also analyzed. Further application are given to hardness\nof two natural problems in Integer Optimizations.\n",
"title": "Short Presburger arithmetic is hard"
}
| null | null | null | null | true | null |
6363
| null |
Default
| null | null |
null |
{
"abstract": " A primary goal of galaxy surveys is to tighten constraints on cosmological\nparameters, and the power spectrum $P(k)$ is the standard means of doing so.\nHowever, at translinear scales $P(k)$ is blind to much of these surveys'\ninformation---information which the log density power spectrum recovers. For\ndiscrete fields (such as the galaxy density), $A^*$ denotes the statistic\nanalogous to the log density: $A^*$ is a \"sufficient statistic\" in that its\npower spectrum (and mean) capture virtually all of a discrete survey's\ninformation. However, the power spectrum of $A^*$ is biased with respect to the\ncorresponding log spectrum for continuous fields, and to use $P_{A^*}(k)$ to\nconstrain the values of cosmological parameters, we require some means of\npredicting this bias. Here we present a prescription for doing so; for\nEuclid-like surveys (with cubical cells 16$h^{-1}$ Mpc across) our bias\nprescription's error is less than 3 per cent. This prediction will facilitate\noptimal utilization of the information in future galaxy surveys.\n",
"title": "The Bias of the Log Power Spectrum for Discrete Surveys"
}
| null | null | null | null | true | null |
6364
| null |
Default
| null | null |
null |
{
"abstract": " A weakly dependent time series regression model with multivariate covariates\nand univariate observations is considered, for which we develop a procedure to\ndetect whether the nonparametric conditional mean function is stable in time\nagainst change point alternatives. Our proposal is based on a modified CUSUM\ntype test procedure, which uses a sequential marked empirical process of\nresiduals. We show weak convergence of the considered process to a centered\nGaussian process under the null hypothesis of no change in the mean function\nand a stationarity assumption. This requires some sophisticated arguments for\nsequential empirical processes of weakly dependent variables. As a consequence\nwe obtain convergence of Kolmogorov-Smirnov and Cramér-von Mises type test\nstatistics. The proposed procedure acquires a very simple limiting distribution\nand nice consistency properties, features from which related tests are lacking.\nWe moreover suggest a bootstrap version of the procedure and discuss its\napplicability in the case of unstable variances.\n",
"title": "Consistent nonparametric change point detection combining CUSUM and marked empirical processes"
}
| null | null | null | null | true | null |
6365
| null |
Default
| null | null |
null |
{
"abstract": " The electric field effect on magnetic anisotropy was studied in an ultrathin\nFe(001) monocrystalline layer sandwiched between Cr buffer and MgO tunnel\nbarrier layers, mainly through post-annealing temperature and measurement\ntemperature dependences. A large coefficient of the electric field effect of\nmore than 200 fJ/Vm was observed in the negative range of electric field, as\nwell as an areal energy density of perpendicular magnetic anisotropy (PMA) of\naround 600 uJ/m2. More interestingly, nonlinear behavior, giving rise to a\nlocal minimum around +100 mV/nm, was observed in the electric field dependence\nof magnetic anisotropy, being independent of the post-annealing and measurement\ntemperatures. The insensitivity to both the interface conditions and the\ntemperature of the system suggests that the nonlinear behavior is attributed to\nan intrinsic origin such as an inherent electronic structure in the Fe/MgO\ninterface. The present study can contribute to the progress in theoretical\nstudies, such as ab initio calculations, on the mechanism of the electric field\neffect on PMA.\n",
"title": "Nonlinear electric field effect on perpendicular magnetic anisotropy in Fe/MgO interfaces"
}
| null | null |
[
"Physics"
] | null | true | null |
6366
| null |
Validated
| null | null |
null |
{
"abstract": " \"Coevolving\" or \"adaptive\" voter models (AVMs) are natural systems for\nmodeling the emergence of mesoscopic structure from local networked processes\ndriven by conflict and homophily. Because of this, many methods for\napproximating the long-run behavior of AVMs have been proposed over the last\ndecade. However, most such methods are either restricted in scope, expensive in\ncomputation, or inaccurate in predicting important statistics. In this work, we\ndevelop a novel, second-order moment closure approximation method for studying\nthe equilibrium mesoscopic structure of AVMs and apply it to binary-state\nrewire-to-random and rewire-to-same model variants with random state-switching.\nThis framework exploits an asymmetry in voting events that enables us to derive\nanalytic approximations for the fast-timescale dynamics. The resulting\nnumerical approximations enable the computation of key properties of the model\nbehavior, such as the location of the fragmentation transition and the\nequilibrium active edge density, across the entire range of state densities.\nNumerically, they are nearly exact for the rewire-to-random model, and\ncompetitive with other current approaches for the rewire-to-same model. We\nconclude with suggestions for model refinement and extensions to more complex\nmodels.\n",
"title": "Local Symmetry and Global Structure in Adaptive Voter Models"
}
| null | null | null | null | true | null |
6367
| null |
Default
| null | null |
null |
{
"abstract": " This article provides a weighted model confidence set, whenever underling\nmodel has been misspecified and some part of support of random variable $X$\nconveys some important information about underling true model. Application of\nsuch weighted model confidence set for local and mixture model confidence sets\nhave been given. Two simulation studies have been conducted to show practical\napplication of our findings.\n",
"title": "A Weighted Model Confidence Set: Applications to Local and Mixture Model Confidence Sets"
}
| null | null | null | null | true | null |
6368
| null |
Default
| null | null |
null |
{
"abstract": " Mapping the spatial distribution of poverty in developing countries remains\nan important and costly challenge. These \"poverty maps\" are key inputs for\npoverty targeting, public goods provision, political accountability, and impact\nevaluation, that are all the more important given the geographic dispersion of\nthe remaining bottom billion severely poor individuals. In this paper we train\nConvolutional Neural Networks (CNNs) to estimate poverty directly from high and\nmedium resolution satellite images. We use both Planet and Digital Globe\nimagery with spatial resolutions of 3-5 sq. m. and 50 sq. cm. respectively,\ncovering all 2 million sq. km. of Mexico. Benchmark poverty estimates come from\nthe 2014 MCS-ENIGH combined with the 2015 Intercensus and are used to estimate\npoverty rates for 2,456 Mexican municipalities. CNNs are trained using the 896\nmunicipalities in the 2014 MCS-ENIGH. We experiment with several architectures\n(GoogleNet, VGG) and use GoogleNet as a final architecture where weights are\nfine-tuned from ImageNet. We find that 1) the best models, which incorporate\nsatellite-estimated land use as a predictor, explain approximately 57% of the\nvariation in poverty in a validation sample of 10 percent of MCS-ENIGH\nmunicipalities; 2) Across all MCS-ENIGH municipalities explanatory power\nreduces to 44% in a CNN prediction and landcover model; 3) Predicted poverty\nfrom the CNN predictions alone explains 47% of the variation in poverty in the\nvalidation sample, and 37% over all MCS-ENIGH municipalities; 4) In urban areas\nwe see slight improvements from using Digital Globe versus Planet imagery,\nwhich explain 61% and 54% of poverty variation respectively. We conclude that\nCNNs can be trained end-to-end on satellite imagery to estimate poverty,\nalthough there is much work to be done to understand how the training process\ninfluences out of sample validation.\n",
"title": "Poverty Mapping Using Convolutional Neural Networks Trained on High and Medium Resolution Satellite Images, With an Application in Mexico"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6369
| null |
Validated
| null | null |
null |
{
"abstract": " The magnetic signature of an urban environment is investigated using a\ngeographically distributed network of fluxgate magnetometers deployed in and\naround Berkeley, California. The system hardware and software are described and\nresults from initial operation of the network are reported. The sensors sample\nthe vector magnetic field with a 4 kHz resolution and are sensitive to\nfluctuations below 0.1 $\\textrm{nT}/\\sqrt{\\textrm{Hz}}$. Data from separate\nstations are synchronized to around $\\pm100$ $\\mu{s}$ using GPS and computer\nsystem clocks. Data from all sensors are automatically uploaded to a central\nserver. Anomalous events, such as lightning strikes, have been observed. A\nwavelet analysis is used to study observations over a wide range of temporal\nscales up to daily variations that show strong differences between weekend and\nweekdays. The Bay Area Rapid Transit (BART) is identified as the most dominant\nsignal from these observations and a superposed epoch analysis is used to study\nand extract the BART signal. Initial results of the correlation between sensors\nare also presented.\n",
"title": "Network of sensitive magnetometers for urban studies"
}
| null | null | null | null | true | null |
6370
| null |
Default
| null | null |
null |
{
"abstract": " The cooperative hierarchical structure is a common and significant data\nstructure observed in, or adopted by, many research areas, such as: text mining\n(author-paper-word) and multi-label classification (label-instance-feature).\nRenowned Bayesian approaches for cooperative hierarchical structure modeling\nare mostly based on topic models. However, these approaches suffer from a\nserious issue in that the number of hidden topics/factors needs to be fixed in\nadvance and an inappropriate number may lead to overfitting or underfitting.\nOne elegant way to resolve this issue is Bayesian nonparametric learning, but\nexisting work in this area still cannot be applied to cooperative hierarchical\nstructure modeling.\nIn this paper, we propose a cooperative hierarchical Dirichlet process (CHDP)\nto fill this gap. Each node in a cooperative hierarchical structure is assigned\na Dirichlet process to model its weights on the infinite hidden factors/topics.\nTogether with measure inheritance from hierarchical Dirichlet process, two\nkinds of measure cooperation, i.e., superposition and maximization, are defined\nto capture the many-to-many relationships in the cooperative hierarchical\nstructure. Furthermore, two constructive representations for CHDP, i.e.,\nstick-breaking and international restaurant process, are designed to facilitate\nthe model inference. Experiments on synthetic and real-world data with\ncooperative hierarchical structures demonstrate the properties and the ability\nof CHDP for cooperative hierarchical structure modeling and its potential for\npractical application scenarios.\n",
"title": "Cooperative Hierarchical Dirichlet Processes: Superposition vs. Maximization"
}
| null | null | null | null | true | null |
6371
| null |
Default
| null | null |
null |
{
"abstract": " We present a model for the origin of the extended law of star formation in\nwhich the surface density of star formation ($\\Sigma_{\\rm SFR}$) depends not\nonly on the local surface density of the gas ($\\Sigma_{g}$), but also on the\nstellar surface density ($\\Sigma_{*}$), the velocity dispersion of the stars,\nand on the scaling laws of turbulence in the gas. We compare our model with the\nspiral, face-on galaxy NGC 628 and show that the dependence of the star\nformation rate on the entire set of physical quantities for both gas and stars\ncan help explain both the observed general trends in the\n$\\Sigma_{g}-\\Sigma_{\\rm SFR}$ and $\\Sigma_{*}-\\Sigma_{\\rm SFR}$ relations, but\nalso, and equally important, the scatter in these relations at any value of\n$\\Sigma_{g}$ and $\\Sigma_{*}$. Our results point out to the crucial role played\nby existing stars along with the gaseous component in setting the conditions\nfor large scale gravitational instabilities and star formation in galactic\ndisks.\n",
"title": "The extended law of star formation: the combined role of gas and stars"
}
| null | null | null | null | true | null |
6372
| null |
Default
| null | null |
null |
{
"abstract": " R. Guralnick (Linear Algebra Appl. 99, 85-96, 1988) proved that two\nholomorphic matrices on a noncompact connected Riemann surface, which are\nlocally holomorphically similar, are globally holomorphically similar. We\ngeneralize this to (possibly, non-smooth) one-dimensional Stein spaces. For\nStein spaces of arbitrary dimension, we prove that global $\\mathcal C^\\infty$\nsimilarity implies global holomorphic similarity, whereas global continuous\nsimilarity is not sufficient.\n",
"title": "Local and global similarity of holomorphic matrices"
}
| null | null | null | null | true | null |
6373
| null |
Default
| null | null |
null |
{
"abstract": " Learning sparse linear models with two-way interactions is desirable in many\napplication domains such as genomics. l1-regularised linear models are popular\nto estimate sparse models, yet standard implementations fail to address\nspecifically the quadratic explosion of candidate two-way interactions in high\ndimensions, and typically do not scale to genetic data with hundreds of\nthousands of features. Here we present WHInter, a working set algorithm to\nsolve large l1-regularised problems with two-way interactions for binary design\nmatrices. The novelty of WHInter stems from a new bound to efficiently identify\nworking sets while avoiding to scan all features, and on fast computations\ninspired from solutions to the maximum inner product search problem. We apply\nWHInter to simulated and real genetic data and show that it is more scalable\nand two orders of magnitude faster than the state of the art.\n",
"title": "WHInter: A Working set algorithm for High-dimensional sparse second order Interaction models"
}
| null | null | null | null | true | null |
6374
| null |
Default
| null | null |
null |
{
"abstract": " Convolutional Neural Networks (CNNs) have achieved state-of-the-art\nperformance on a variety of computer vision tasks, particularly visual\nclassification problems, where new algorithms reported to achieve or even\nsurpass the human performance. In this paper, we examine whether CNNs are\ncapable of learning the semantics of training data. To this end, we evaluate\nCNNs on negative images, since they share the same structure and semantics as\nregular images and humans can classify them correctly. Our experimental results\nindicate that when training on regular images and testing on negative images,\nthe model accuracy is significantly lower than when it is tested on regular\nimages. This leads us to the conjecture that current training methods do not\neffectively train models to generalize the concepts. We then introduce the\nnotion of semantic adversarial examples - transformed inputs that semantically\nrepresent the same objects, but the model does not classify them correctly -\nand present negative images as one class of such inputs.\n",
"title": "On the Limitation of Convolutional Neural Networks in Recognizing Negative Images"
}
| null | null | null | null | true | null |
6375
| null |
Default
| null | null |
null |
{
"abstract": " For a regular cardinal $\\kappa$, a formula of the modal $\\mu$-calculus is\n$\\kappa$-continuous in a variable x if, on every model, its interpretation as a\nunary function of x is monotone and preserves unions of $\\kappa$-directed sets.\nWe define the fragment $C_{\\aleph_1}(x)$ of the modal $\\mu$-calculus and prove\nthat all the formulas in this fragment are $\\aleph_1$-continuous. For each\nformula $\\phi(x)$ of the modal $\\mu$-calculus, we construct a formula $\\psi(x)\n\\in C_{\\aleph_1 }(x)$ such that $\\phi(x)$ is $\\kappa$-continuous, for some\n$\\kappa$, if and only if $\\phi(x)$ is equivalent to $\\psi(x)$. Consequently, we\nprove that (i) the problem whether a formula is $\\kappa$-continuous for some\n$\\kappa$ is decidable, (ii) up to equivalence, there are only two fragments\ndetermined by continuity at some regular cardinal: the fragment\n$C_{\\aleph_0}(x)$ studied by Fontaine and the fragment $C_{\\aleph_1}(x)$. We\napply our considerations to the problem of characterizing closure ordinals of\nformulas of the modal $\\mu$-calculus. An ordinal $\\alpha$ is the closure\nordinal of a formula $\\phi(x)$ if its interpretation on every model converges\nto its least fixed-point in at most $\\alpha$ steps and if there is a model\nwhere the convergence occurs exactly in $\\alpha$ steps. We prove that\n$\\omega_1$, the least uncountable ordinal, is such a closure ordinal. Moreover\nwe prove that closure ordinals are closed under ordinal sum. Thus, any formal\nexpression built from 0, 1, $\\omega$, $\\omega_1$ by using the binary operator\nsymbol + gives rise to a closure ordinal.\n",
"title": "$\\aleph_1$ and the modal $μ$-calculus"
}
| null | null | null | null | true | null |
6376
| null |
Default
| null | null |
null |
{
"abstract": " A fundamental challenge in large-scale cloud networks and data centers is to\nachieve highly efficient server utilization and limit energy consumption, while\nproviding excellent user-perceived performance in the presence of uncertain and\ntime-varying demand patterns. Auto-scaling provides a popular paradigm for\nautomatically adjusting service capacity in response to demand while meeting\nperformance targets, and queue-driven auto-scaling techniques have been widely\ninvestigated in the literature. In typical data center architectures and cloud\nenvironments however, no centralized queue is maintained, and load balancing\nalgorithms immediately distribute incoming tasks among parallel queues. In\nthese distributed settings with vast numbers of servers, centralized\nqueue-driven auto-scaling techniques involve a substantial communication\noverhead and major implementation burden, or may not even be viable at all.\nMotivated by the above issues, we propose a joint auto-scaling and load\nbalancing scheme which does not require any global queue length information or\nexplicit knowledge of system parameters, and yet provides provably near-optimal\nservice elasticity. We establish the fluid-level dynamics for the proposed\nscheme in a regime where the total traffic volume and nominal service capacity\ngrow large in proportion. The fluid-limit results show that the proposed scheme\nachieves asymptotic optimality in terms of user-perceived delay performance as\nwell as energy consumption. Specifically, we prove that both the waiting time\nof tasks and the relative energy portion consumed by idle servers vanish in the\nlimit. At the same time, the proposed scheme operates in a distributed fashion\nand involves only constant communication overhead per task, thus ensuring\nscalability in massive data center operations.\n",
"title": "Optimal Service Elasticity in Large-Scale Distributed Systems"
}
| null | null | null | null | true | null |
6377
| null |
Default
| null | null |
null |
{
"abstract": " The hour-glass-like dispersion of spin excitations is a common feature of\nunderdoped cuprates. It was qualitatively explained by the random phase\napproximation based on various ordered states with some phenomenological\nparameters; however, its origin remains elusive. Here, we present a numerical\nstudy of spin dynamics in the $t$-$J$ model using the variational Monte Carlo\nmethod. This parameter-free method satisfies the no double-occupancy constraint\nof the model and thus provides a better evaluation on the spin dynamics with\nrespect to various mean-field trial states. We conclude that the lower branch\nof the hour-glass dispersion is a collective mode and the upper branch is more\nlikely the consequence of the stripe state than the other candidates.\n",
"title": "Variational Monte Carlo study of spin dynamics in underdoped cuprates"
}
| null | null | null | null | true | null |
6378
| null |
Default
| null | null |
null |
{
"abstract": " I propose to use high brightness electron beam with 1 to 100 MeV energy as\ntool to combat tumor or cancerous tissues in deep part of body. The method is\nto directly deliver the electron beam to the tumor site via a small tube that\nconnected to a high brightness electron beam accelerator that is commonly\navailable around the world. Here I gave a basic scheme on the principle, I\nbelieve other issues people raises will be solved easily for those who are\ninterested in solving the problems.\n",
"title": "High brightness electron beam for radiation therapy: A new approach"
}
| null | null | null | null | true | null |
6379
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study the fundamental solution $\\varGamma(t,x;\\tau,\\xi)$ of\nthe parabolic operator $L_{t}=\\partial_{t}-\\Delta+b(t,x)\\cdot\\nabla$, where for\nevery $t$, $b(t,\\cdot)$ is a divergence-free vector field, and we consider the\ncase that $b$ belongs to the Lebesgue space\n$L^{l}\\left(0,T;L^{q}\\left(\\mathbb{R}^{n}\\right)\\right)$. The regularity of\nweak solutions to the parabolic equation $L_{t}u=0$ depends critically on the\nvalue of the parabolic exponent $\\gamma=\\frac{2}{l}+\\frac{n}{q}$. Without the\ndivergence-free condition on $b$, the regularity of weak solutions has been\nestablished when $\\gamma\\leq1$, and the heat kernel estimate has been obtained\nas well, except for the case that $l=\\infty,q=n$. The regularity of weak\nsolutions was deemed not true for the critical case\n$L^{\\infty}\\left(0,T;L^{n}\\left(\\mathbb{R}^{n}\\right)\\right)$ for a general\n$b$, while it is true for the divergence-free case, and a written proof can be\ndeduced from the results in [Semenov, 2006]. One of the results obtained in the\npresent paper establishes the Aronson type estimate for critical and\nsupercritical cases and for vector fields $b$ which are divergence-free. We\nwill prove the best possible lower and upper bounds for the fundamental\nsolution one can derive under the current approach. The significance of the\ndivergence-free condition enters the study of parabolic equations rather\nrecently, mainly due to the discovery of the compensated compactness. The\ninterest for the study of such parabolic equations comes from its connections\nwith Leray's weak solutions of the Navier-Stokes equations and the Taylor\ndiffusion associated with a vector field where the heat operator $L_{t}$\nappears naturally.\n",
"title": "Parabolic equations with divergence-free drift in space $L_{t}^{l}L_{x}^{q}$"
}
| null | null | null | null | true | null |
6380
| null |
Default
| null | null |
null |
{
"abstract": " Our work presented in this paper focuses on the translation of terminological\nexpressions represented in semantically structured resources, like ontologies\nor knowledge graphs. The challenge of translating ontology labels or\nterminological expressions represented in knowledge bases lies in the highly\nspecific vocabulary and the lack of contextual information, which can guide a\nmachine translation system to translate ambiguous words into the targeted\ndomain. Due to these challenges, we evaluate the translation quality of\ndomain-specific expressions in the medical and financial domain with\nstatistical (SMT) as well as with neural machine translation (NMT) methods and\nexperiment domain adaptation of the translation models with terminological\nexpressions only. Furthermore, we perform experiments on the injection of\nexternal terminological expressions into the translation systems. Through these\nexperiments, we observed a significant advantage in domain adaptation for the\ndomain-specific resource in the medical and financial domain and the benefit of\nsubword models over word-based NMT models for terminology translation.\n",
"title": "Translating Terminological Expressions in Knowledge Bases with Neural Machine Translation"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6381
| null |
Validated
| null | null |
null |
{
"abstract": " This paper focuses on Byzantine attack detection for Gaussian two-hop one-way\nrelay network, where an amplify-and-forward relay may conduct Byzantine attacks\nby forwarding altered symbols to the destination. For facilitating attack\ndetection, we utilize the openness of wireless medium to make the destination\nobserve some secured signals that are not attacked. Then, a detection scheme is\ndeveloped for the destination by using its secured observations to\nstatistically check other observations from the relay. On the other hand,\nnotice the Gaussian channel is continuous, which allows the possible Byzantine\nattacks to be conducted within continuous alphabet(s). The existing work on\ndiscrete channel is not applicable for investigating the performance of the\nproposed scheme. The main contribution of this paper is to prove that if and\nonly if the wireless relay network satisfies a non-manipulable channel\ncondition, the proposed detection scheme achieves asymptotic errorless\nperformance against arbitrary attacks that allow the stochastic distributions\nof altered symbols to vary arbitrarily and depend on each other. No pre-shared\nsecret or secret transmission is needed for the detection. Furthermore, we also\nprove that the relay network is non-manipulable as long as all channel\ncoefficients are non-zero, which is not essential restrict for many practical\nsystems.\n",
"title": "Detecting Arbitrary Attacks Using Continuous Secured Side Information in Wireless Networks"
}
| null | null | null | null | true | null |
6382
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the pooled data problem of identifying the labels\nassociated with a large collection of items, based on a sequence of pooled\ntests revealing the counts of each label within the pool. In the noiseless\nsetting, we identify an exact asymptotic threshold on the required number of\ntests with optimal decoding, and prove a phase transition between complete\nsuccess and complete failure. In addition, we present a novel noisy variation\nof the problem, and provide an information-theoretic framework for\ncharacterizing the required number of tests for general random noise models.\nOur results reveal that noise can make the problem considerably more difficult,\nwith strict increases in the scaling laws even at low noise levels. Finally, we\ndemonstrate similar behavior in an approximate recovery setting, where a given\nnumber of errors is allowed in the decoded labels.\n",
"title": "Phase Transitions in the Pooled Data Problem"
}
| null | null | null | null | true | null |
6383
| null |
Default
| null | null |
null |
{
"abstract": " It is a usual practice to ignore any structural information underlying\nclasses in multi-class classification. In this paper, we propose a graph\nconvolutional network (GCN) augmented neural network classifier to exploit a\nknown, underlying graph structure of labels. The proposed approach resembles an\n(approximate) inference procedure in, for instance, a conditional random field\n(CRF). We evaluate the proposed approach on document classification and object\nrecognition and report both accuracies and graph-theoretic metrics that\ncorrespond to the consistency of the model's prediction. The experiment results\nreveal that the proposed model outperforms a baseline method which ignores the\ngraph structures of a label space in terms of graph-theoretic metrics.\n",
"title": "Graph Convolutional Networks for Classification with a Structured Label Space"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6384
| null |
Validated
| null | null |
null |
{
"abstract": " The combination of the surface science techniques (STM, XPS, ARPES) and\ndensity-functional theory calculations was used to study the decoupling of\ngraphene from Ni(111) by oxygen intercalation. The formation of the\nantiferromagnetic (AFM) NiO layer at the interface between graphene and\nferromagnetic (FM) Ni is found, where graphene protects the underlying AFM/FM\nsandwich system. It is found that graphene is fully decoupled in this system\nand strongly $p$-doped via charge transfer with a position of the Dirac point\nof $(0.69\\pm0.02)$ eV above the Fermi level. Our theoretical analysis confirms\nall experimental findings, addressing also the interface properties between\ngraphene and AFM NiO.\n",
"title": "Decoupling of graphene from Ni(111) via oxygen intercalation"
}
| null | null |
[
"Physics"
] | null | true | null |
6385
| null |
Validated
| null | null |
null |
{
"abstract": " The paper proposes a new approach to model risk measurement based on the\nWasserstein distance between two probability measures. It formulates the\ntheoretical motivation resulting from the interpretation of fictitious\nadversary of robust risk management. The proposed approach accounts for all\nalternative models and incorporates the economic reality of the fictitious\nadversary. It provides practically feasible results that overcome the\nrestriction and the integrability issue imposed by the nominal model. The\nWasserstein approach suits for all types of model risk problems, ranging from\nthe single-asset hedging risk problem to the multi-asset allocation problem.\nThe robust capital allocation line, accounting for the correlation risk, is not\nachievable with other non-parametric approaches.\n",
"title": "Model Risk Measurement under Wasserstein Distance"
}
| null | null | null | null | true | null |
6386
| null |
Default
| null | null |
null |
{
"abstract": " A significantly faster algorithm is presented for the original kNN mode\nseeking procedure. It has the advantages over the well-known mean shift\nalgorithm that it is feasible in high-dimensional vector spaces and results in\nuniquely, well defined modes. Moreover, without any additional computational\neffort it may yield a multi-scale hierarchy of clusterings. The time complexity\nis just O(n^1.5). resulting computing times range from seconds for 10^4 objects\nto minutes for 10^5 objects and to less than an hour for 10^6 objects. The\nspace complexity is just O(n). The procedure is well suited for finding large\nsets of small clusters and is thereby a candidate to analyze thousands of\nclusters in millions of objects.\nThe kNN mode seeking procedure can be used for active learning by assigning\nthe clusters to the class of the modal objects of the clusters. Its feasibility\nis shown by some examples with up to 1.5 million handwritten digits. The\nobtained classification results based on the clusterings are compared with\nthose obtained by the nearest neighbor rule and the support vector classifier\nbased on the same labeled objects for training. It can be concluded that using\nthe clustering structure for classification can be significantly better than\nusing the trained classifiers. A drawback of using the clustering for\nclassification, however, is that no classifier is obtained that may be used for\nout-of-sample objects.\n",
"title": "Fast kNN mode seeking clustering applied to active learning"
}
| null | null | null | null | true | null |
6387
| null |
Default
| null | null |
null |
{
"abstract": " This paper investigates a flow- and path-sensitive static information flow\nanalysis. Compared with security type systems with fixed labels, it has been\nshown that flow-sensitive type systems accept more secure programs. We show\nthat an information flow analysis with fixed labels can be both flow- and\npath-sensitive. The novel analysis has two major components: 1) a\ngeneral-purpose program transformation that removes false dataflow dependencies\nin a program that confuse a fixed-label type system, and 2) a fixed-label type\nsystem that allows security types to depend on path conditions. We formally\nprove that the proposed analysis enforces a rigorous security property:\nnoninterference. Moreover, we show that the analysis is strictly more precise\nthan a classic flow-sensitive type system, and it allows sound control of\ninformation flow in the presence of mutable variables without resorting to\nrun-time mechanisms.\n",
"title": "Towards a Flow- and Path-Sensitive Information Flow Analysis: Technical Report"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6388
| null |
Validated
| null | null |
null |
{
"abstract": " Topic models have been extensively used to organize and interpret the\ncontents of large, unstructured corpora of text documents. Although topic\nmodels often perform well on traditional training vs. test set evaluations, it\nis often the case that the results of a topic model do not align with human\ninterpretation. This interpretability fallacy is largely due to the\nunsupervised nature of topic models, which prohibits any user guidance on the\nresults of a model. In this paper, we introduce a semi-supervised method called\ntopic supervised non-negative matrix factorization (TS-NMF) that enables the\nuser to provide labeled example documents to promote the discovery of more\nmeaningful semantic structure of a corpus. In this way, the results of TS-NMF\nbetter match the intuition and desired labeling of the user. The core of TS-NMF\nrelies on solving a non-convex optimization problem for which we derive an\niterative algorithm that is shown to be monotonic and convergent to a local\noptimum. We demonstrate the practical utility of TS-NMF on the Reuters and\nPubMed corpora, and find that TS-NMF is especially useful for conceptual or\nbroad topics, where topic key terms are not well understood. Although\nidentifying an optimal latent structure for the data is not a primary objective\nof the proposed approach, we find that TS-NMF achieves higher weighted Jaccard\nsimilarity scores than the contemporary methods, (unsupervised) NMF and latent\nDirichlet allocation, at supervision rates as low as 10% to 20%.\n",
"title": "Topic supervised non-negative matrix factorization"
}
| null | null | null | null | true | null |
6389
| null |
Default
| null | null |
null |
{
"abstract": " The quasi-two-dimensional organic charge-transfer salt\n$\\kappa$-(BEDT-TTF)$_2$Cu$_2$(CN)$_3$ is one of the prime candidates for a\nquantum spin-liquid due the strong spin frustration of its anisotropic\ntriangular lattice in combination with its proximity to the Mott transition.\nDespite intensive investigations of the material's low-temperature properties,\nseveral important questions remain to be answered. Particularly puzzling are\nthe 6\\,K anomaly and the enigmatic effects observed in magnetic fields. Here we\nreport on low-temperature measurements of lattice effects which were shown to\nbe particularly strongly pronounced in this material (R. S. Manna \\emph{et\nal.}, Phys. Rev. Lett. \\textbf{104}, 016403 (2010)). A special focus of our\nstudy lies on sample-to-sample variations of these effects and their\nimplications on the interpretation of experimental data. By investigating\noverall nine single crystals from two different batches, we can state that\nthere are considerable differences in the size of the second-order phase\ntransition anomaly around 6\\,K, varying within a factor of 3. In addition, we\nfind field-induced anomalies giving rise to pronounced features in the sample\nlength for two out of these nine crystals for temperatures $T <$ 9 K. We\ntentatively assign the latter effects to $B$-induced magnetic clusters\nsuspected to nucleate around crystal imperfections. These $B$-induced effects\nare absent for the crystals where the 6\\,K anomaly is most strongly pronounced.\nThe large lattice effects observed at 6\\,K are consistent with proposed pairing\ninstabilities of fermionic excitations breaking the lattice symmetry. The\nstrong sample-to-sample variation in the size of the phase transition anomaly\nsuggests that the conversion of the fermions to bosons at the instability is\nonly partial and to some extent influenced by not yet identified\nsample-specific parameters.\n",
"title": "Low-temperature lattice effects in the spin-liquid candidate $κ$-(BEDT-TTF)$_2$Cu$_2$(CN)$_3$"
}
| null | null |
[
"Physics"
] | null | true | null |
6390
| null |
Validated
| null | null |
null |
{
"abstract": " With advanced data analytical techniques, efforts for more accurate decision\nsupport systems for disease prediction are on rise. Surveys by World Health\nOrganization (WHO) indicate a great increase in number of diabetic patients and\nrelated deaths each year. Early diagnosis of diabetes is a major concern among\nresearchers and practitioners. The paper presents an application of\n\\textit{Automatic Multilayer Perceptron }which\\textit{ }is combined with an\noutlier detection method \\textit{Enhanced Class Outlier Detection using\ndistance based algorithm }to create a prediction framework named as Enhanced\nClass Outlier with Automatic Multi layer Perceptron (ECO-AMLP). A series of\nexperiments are performed on publicly available Pima Indian Diabetes Dataset to\ncompare ECO-AMLP with other individual classifiers as well as ensemble based\nmethods. The outlier technique used in our framework gave better results as\ncompared to other pre-processing and classification techniques. Finally, the\nresults are compared with other state-of-the-art methods reported in literature\nfor diabetes prediction on PIDD and achieved accuracy of 88.7\\% bests all other\nreported studies.\n",
"title": "ECO-AMLP: A Decision Support System using an Enhanced Class Outlier with Automatic Multilayer Perceptron for Diabetes Prediction"
}
| null | null | null | null | true | null |
6391
| null |
Default
| null | null |
null |
{
"abstract": " We provide a local approximation result of non-holomorphic discs with small\nd-bar by pseudoholomorphic ones. As an application, we provide a certain gluing\nconstruction.\n",
"title": "Local approximation of non-holomorphic discs in almost complex manifolds"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6392
| null |
Validated
| null | null |
null |
{
"abstract": " Following their success in Computer Vision and other areas, deep learning\ntechniques have recently become widely adopted in Music Information Retrieval\n(MIR) research. However, the majority of works aim to adopt and assess methods\nthat have been shown to be effective in other domains, while there is still a\ngreat need for more original research focusing on music primarily and utilising\nmusical knowledge and insight. The goal of this paper is to boost the interest\nof beginners by providing a comprehensive tutorial and reducing the barriers to\nentry into deep learning for MIR. We lay out the basic principles and review\nprominent works in this hard to navigate the field. We then outline the network\nstructures that have been successful in MIR problems and facilitate the\nselection of building blocks for the problems at hand. Finally, guidelines for\nnew tasks and some advanced topics in deep learning are discussed to stimulate\nnew research in this fascinating field.\n",
"title": "A Tutorial on Deep Learning for Music Information Retrieval"
}
| null | null | null | null | true | null |
6393
| null |
Default
| null | null |
null |
{
"abstract": " This chapter provides an introduction to the modeling and control of power\ngeneration from wind turbine systems. In modeling, the focus is on the\nelectrical components: electrical machine (e.g. permanent-magnet synchronous\ngenerators), back-to-back converter (consisting of machine-side and grid-side\nconverter sharing a common DC-link), mains filters and ideal (balanced) power\ngrid. The aerodynamics and the torque generation of the wind turbine are\nexplained in simplified terms using a so-called power coefficient. The overall\ncontrol system is considered. In particular, the phase-locked loop system for\ngrid-side voltage orientation, the nonlinear speed control system for the\ngenerator (and turbine), and the non-minimum phase DC-link voltage control\nsystem are discussed in detail; based on a brief derivation of the underlying\nmachine-side and grid-side current control systems. With the help of the power\nbalance of the wind turbine, the operation management and the control of the\npower flow are explained. Concluding simulation results illustrate the overall\nsystem behavior of a controlled wind turbine with a permanent-magnet\nsynchronous generator.\n",
"title": "Modeling and control of modern wind turbine systems: An introduction"
}
| null | null | null | null | true | null |
6394
| null |
Default
| null | null |
null |
{
"abstract": " A classical difficult isomorphism testing problem is to test isomorphism of\np-groups of class 2 and exponent p in time polynomial in the group order. It is\nknown that this problem can be reduced to solving the alternating matrix space\nisometry problem over a finite field in time polynomial in the underlying\nvector space size. We propose a venue of attack for the latter problem by\nviewing it as a linear algebraic analogue of the graph isomorphism problem.\nThis viewpoint leads us to explore the possibility of transferring techniques\nfor graph isomorphism to this long-believed bottleneck case of group\nisomorphism.\nIn 1970's, Babai, Erdős, and Selkow presented the first average-case\nefficient graph isomorphism testing algorithm (SIAM J Computing, 1980).\nInspired by that algorithm, we devise an average-case efficient algorithm for\nthe alternating matrix space isometry problem over a key range of parameters,\nin a random model of alternating matrix spaces in vein of the Erdős-Rényi\nmodel of random graphs. For this, we develop a linear algebraic analogue of the\nclassical individualisation technique, a technique belonging to a set of\ncombinatorial techniques that has been critical for the progress on the\nworst-case time complexity for graph isomorphism, but was missing in the group\nisomorphism context. As a consequence of the main algorithm, we establish a\nweaker linear algebraic analogue of Erdős and Rényi's classical result\nthat most graphs have the trivial automorphism group. We finally show that\nLuks' dynamic programming technique for graph isomorphism (STOC 1999) can be\nadapted to slightly improve the worst-case time complexity of the alternating\nmatrix space isometry problem in a certain range of parameters.\n",
"title": "Linear algebraic analogues of the graph isomorphism problem and the Erdős-Rényi model"
}
| null | null | null | null | true | null |
6395
| null |
Default
| null | null |
null |
{
"abstract": " We give the motivation for scoring clustering algorithms and a metric $M : A\n\\rightarrow \\mathbb{N}$ from the set of clustering algorithms to the natural\nnumbers which we realize as \\begin{equation} M(A) = \\sum_i \\alpha_i |f_i -\n\\beta_i|^{w_i} \\end{equation} where $\\alpha_i,\\beta_i,w_i$ are parameters used\nfor scoring the feature $f_i$, which is computed empirically.. We give a method\nby which one can score features such as stability, noise sensitivity, etc and\nderive the necessary parameters. We conclude by giving a sample set of scores.\n",
"title": "A Family of Metrics for Clustering Algorithms"
}
| null | null | null | null | true | null |
6396
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we introduce a new reformulation of the Green-Naghdi model in\nthe Camassa-Holm regime for the propagation of internal waves over a flat\ntopography derived by Duchêne, Israwi and Talhouk. These new Green-Naghdi\nsystems are adapted to improve the frequency dispersion of the original model,\nthey share the same order of precision as the standard one but have an\nappropriate structure which makes them much more suitable for the numerical\nresolution. We develop a second order splitting scheme where the hyperbolic\npart of the system is treated with a high-order finite volume scheme and the\ndispersive part is treated with a finite difference approach. Numerical\nsimulations are then performed to validate the model and the numerical methods.\n",
"title": "A numerical scheme for an improved Green-Naghdi model in the Camassa-Holm regime for the propagation of internal waves"
}
| null | null | null | null | true | null |
6397
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we revisit the weighted likelihood bootstrap, a method that\ngenerates samples from an approximate Bayesian posterior of a parametric model.\nWe show that the same method can be derived, without approximation, under a\nBayesian nonparametric model with the parameter of interest defined as\nminimising an expected negative log-likelihood under an unknown sampling\ndistribution. This interpretation enables us to extend the weighted likelihood\nbootstrap to posterior sampling for parameters minimizing an expected loss. We\ncall this method the loss-likelihood bootstrap. We make a connection between\nthis and general Bayesian updating, which is a way of updating prior belief\ndistributions without needing to construct a global probability model, yet\nrequires the calibration of two forms of loss function. The loss-likelihood\nbootstrap is used to calibrate the general Bayesian posterior by matching\nasymptotic Fisher information. We demonstrate the methodology on a number of\nexamples.\n",
"title": "General Bayesian Updating and the Loss-Likelihood Bootstrap"
}
| null | null | null | null | true | null |
6398
| null |
Default
| null | null |
null |
{
"abstract": " For many applications, an ensemble of base classifiers is an effective\nsolution. The tuning of its parameters(number of classes, amount of data on\nwhich each classifier is to be trained on, etc.) requires G, the generalization\nerror of a given ensemble. The efficient estimation of G is the focus of this\npaper. The key idea is to approximate the variance of the class\nscores/probabilities of the base classifiers over the randomness imposed by the\ntraining subset by normal/beta distribution at each point x in the input\nfeature space. We estimate the parameters of the distribution using a small set\nof randomly chosen base classifiers and use those parameters to give efficient\nestimation schemes for G. We give empirical evidence for the quality of the\nvarious estimators. We also demonstrate their usefulness in making design\nchoices such as the number of classifiers in the ensemble and the size of a\nsubset of data used for training that is needed to achieve a certain value of\ngeneralization error. Our approach also has great potential for designing\ndistributed ensemble classifiers.\n",
"title": "Efficient Estimation of Generalization Error and Bias-Variance Components of Ensembles"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6399
| null |
Validated
| null | null |
null |
{
"abstract": " In the framework of Keldysh-Usadel kinetic theory, we study the temperature\ndependence of flux-flow conductivity (FFC) in diffusive superconductors. By\nusing self-consistent vortex solutions we find the exact values of\ndimensionless parameters that determine the diffusion-controlled FFC both in\nthe limit of the low temperatures and close to the critical one. Taking into\naccount the electron-phonon scattering we study the transition between\nflux-flow regimes controlled either by the diffusion or the inelastic\nrelaxation of non-equilibrium quasiparticles. We demonstrate that the inelastic\nelectron-phonon relaxation leads to the strong suppression of FFC as compared\nto the previous estimates making it possible to obtain the numerical agreement\nwith experimental results.\n",
"title": "Self-consistent calculation of the flux-flow conductivity in diffusive superconductors"
}
| null | null | null | null | true | null |
6400
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.