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": " An empirical investigation of active/continuous authentication for\nsmartphones is presented in this paper by exploiting users' unique application\nusage data, i.e., distinct patterns of use, modeled by a Markovian process.\nVariations of Hidden Markov Models (HMMs) are evaluated for continuous user\nverification, and challenges due to the sparsity of session-wise data, an\nexplosion of states, and handling unforeseen events in the test data are\ntackled. Unlike traditional approaches, the proposed formulation does not\ndepend on the top N-apps, rather uses the complete app-usage information to\nachieve low latency. Through experimentation, empirical assessment of the\nimpact of unforeseen events, i.e., unknown applications and unforeseen\nobservations, on user verification is done via a modified edit-distance\nalgorithm for simple sequence matching. It is found that for enhanced\nverification performance, unforeseen events should be incorporated in the\nmodels by adopting smoothing techniques with HMMs. For validation, extensive\nexperiments on two distinct datasets are performed. The marginal smoothing\ntechnique is the most effective for user verification in terms of equal error\nrate (EER) and with a sampling rate of 1/30s^{-1} and 30 minutes of historical\ndata, and the method is capable of detecting an intrusion within ~2.5 minutes\nof application use.\n",
"title": "Continuous Authentication of Smartphones Based on Application Usage"
}
| null | null | null | null | true | null |
9101
| null |
Default
| null | null |
null |
{
"abstract": " Constraining linear layers in neural networks to respect symmetry\ntransformations from a group $G$ is a common design principle for invariant\nnetworks that has found many applications in machine learning.\nIn this paper, we consider a fundamental question that has received little\nattention to date: Can these networks approximate any (continuous) invariant\nfunction?\nWe tackle the rather general case where $G\\leq S_n$ (an arbitrary subgroup of\nthe symmetric group) that acts on $\\mathbb{R}^n$ by permuting coordinates. This\nsetting includes several recent popular invariant networks. We present two main\nresults: First, $G$-invariant networks are universal if high-order tensors are\nallowed. Second, there are groups $G$ for which higher-order tensors are\nunavoidable for obtaining universality.\n$G$-invariant networks consisting of only first-order tensors are of special\ninterest due to their practical value. We conclude the paper by proving a\nnecessary condition for the universality of $G$-invariant networks that\nincorporate only first-order tensors. Lastly, we propose a conjecture stating\nthat this condition is also sufficient.\n",
"title": "On the Universality of Invariant Networks"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9102
| null |
Validated
| null | null |
null |
{
"abstract": " We study how the behavior of deep policy gradient algorithms reflects the\nconceptual framework motivating their development. We propose a fine-grained\nanalysis of state-of-the-art methods based on key aspects of this framework:\ngradient estimation, value prediction, optimization landscapes, and trust\nregion enforcement. We find that from this perspective, the behavior of deep\npolicy gradient algorithms often deviates from what their motivating framework\nwould predict. Our analysis suggests first steps towards solidifying the\nfoundations of these algorithms, and in particular indicates that we may need\nto move beyond the current benchmark-centric evaluation methodology.\n",
"title": "Are Deep Policy Gradient Algorithms Truly Policy Gradient Algorithms?"
}
| null | null | null | null | true | null |
9103
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we improve the previously best known regret bound to achieve\n$\\epsilon$-differential privacy in oblivious adversarial bandits from\n$\\mathcal{O}{(T^{2/3}/\\epsilon)}$ to $\\mathcal{O}{(\\sqrt{T} \\ln T /\\epsilon)}$.\nThis is achieved by combining a Laplace Mechanism with EXP3. We show that\nthough EXP3 is already differentially private, it leaks a linear amount of\ninformation in $T$. However, we can improve this privacy by relying on its\nintrinsic exponential mechanism for selecting actions. This allows us to reach\n$\\mathcal{O}{(\\sqrt{\\ln T})}$-DP, with a regret of $\\mathcal{O}{(T^{2/3})}$\nthat holds against an adaptive adversary, an improvement from the best known of\n$\\mathcal{O}{(T^{3/4})}$. This is done by using an algorithm that run EXP3 in a\nmini-batch loop. Finally, we run experiments that clearly demonstrate the\nvalidity of our theoretical analysis.\n",
"title": "Achieving Privacy in the Adversarial Multi-Armed Bandit"
}
| null | null | null | null | true | null |
9104
| null |
Default
| null | null |
null |
{
"abstract": " Variational inference is a powerful approach for approximate posterior\ninference. However, it is sensitive to initialization and can be subject to\npoor local optima. In this paper, we develop proximity variational inference\n(PVI). PVI is a new method for optimizing the variational objective that\nconstrains subsequent iterates of the variational parameters to robustify the\noptimization path. Consequently, PVI is less sensitive to initialization and\noptimization quirks and finds better local optima. We demonstrate our method on\nthree proximity statistics. We study PVI on a Bernoulli factor model and\nsigmoid belief network with both real and synthetic data and compare to\ndeterministic annealing (Katahira et al., 2008). We highlight the flexibility\nof PVI by designing a proximity statistic for Bayesian deep learning models\nsuch as the variational autoencoder (Kingma and Welling, 2014; Rezende et al.,\n2014). Empirically, we show that PVI consistently finds better local optima and\ngives better predictive performance.\n",
"title": "Proximity Variational Inference"
}
| null | null | null | null | true | null |
9105
| null |
Default
| null | null |
null |
{
"abstract": " We propose an Encoder-Classifier framework to model the Mandarin tones using\nrecurrent neural networks (RNN). In this framework, extracted frames of\nfeatures for tone classification are fed in to the RNN and casted into a fixed\ndimensional vector (tone embedding) and then classified into tone types using a\nsoftmax layer along with other auxiliary inputs. We investigate various\nconfigurations that help to improve the model, including pooling, feature\nsplicing and utilization of syllable-level tone embeddings. Besides, tone\nembeddings and durations of the contextual syllables are exploited to\nfacilitate tone classification. Experimental results on Mandarin tone\nclassification show the proposed network setups improve tone classification\naccuracy. The results indicate that the RNN encoder-classifier based tone model\nflexibly accommodates heterogeneous inputs (sequential and segmental) and hence\nhas the advantages from both the sequential classification tone models and\nsegmental classification tone models.\n",
"title": "Mandarin tone modeling using recurrent neural networks"
}
| null | null | null | null | true | null |
9106
| null |
Default
| null | null |
null |
{
"abstract": " We modify the nonlinear shallow water equations, the Korteweg-de Vries\nequation, and the Whitham equation, to permit constant vorticity, and examine\nwave breaking, or the lack thereof. By wave breaking, we mean that the solution\nremains bounded but its slope becomes unbounded in finite time. We show that a\nsolution of the vorticity-modified shallow water equations breaks if it carries\nan increase of elevation; the breaking time decreases to zero as the size of\nvorticity increases. We propose a full-dispersion shallow water model, which\ncombines the dispersion relation of water waves and the nonlinear shallow water\nequations in the constant vorticity setting, and which extends the Whitham\nequation to permit bidirectional propagation. We show that its small amplitude\nand periodic traveling wave is unstable to long wavelength perturbations if the\nwave number is greater than a critical value, and stable otherwise, similarly\nto the Benjamin-Feir instability in the irrotational setting; the critical wave\nnumber grows unboundedly large with the size of vorticity. The result agrees\nwith that from a multiple scale expansion of the physical problem. We show that\nvorticity considerably alters the modulational stability and instability in the\npresence of the effects of surface tension.\n",
"title": "Shallow water models with constant vorticity"
}
| null | null | null | null | true | null |
9107
| null |
Default
| null | null |
null |
{
"abstract": " We solve tensor balancing, rescaling an Nth order nonnegative tensor by\nmultiplying N tensors of order N - 1 so that every fiber sums to one. This\ngeneralizes a fundamental process of matrix balancing used to compare matrices\nin a wide range of applications from biology to economics. We present an\nefficient balancing algorithm with quadratic convergence using Newton's method\nand show in numerical experiments that the proposed algorithm is several orders\nof magnitude faster than existing ones. To theoretically prove the correctness\nof the algorithm, we model tensors as probability distributions in a\nstatistical manifold and realize tensor balancing as projection onto a\nsubmanifold. The key to our algorithm is that the gradient of the manifold,\nused as a Jacobian matrix in Newton's method, can be analytically obtained\nusing the Moebius inversion formula, the essential of combinatorial\nmathematics. Our model is not limited to tensor balancing, but has a wide\napplicability as it includes various statistical and machine learning models\nsuch as weighted DAGs and Boltzmann machines.\n",
"title": "Tensor Balancing on Statistical Manifold"
}
| null | null | null | null | true | null |
9108
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of learning an unknown Markov Decision Process (MDP)\nthat is weakly communicating in the infinite horizon setting. We propose a\nThompson Sampling-based reinforcement learning algorithm with dynamic episodes\n(TSDE). At the beginning of each episode, the algorithm generates a sample from\nthe posterior distribution over the unknown model parameters. It then follows\nthe optimal stationary policy for the sampled model for the rest of the\nepisode. The duration of each episode is dynamically determined by two stopping\ncriteria. The first stopping criterion controls the growth rate of episode\nlength. The second stopping criterion happens when the number of visits to any\nstate-action pair is doubled. We establish $\\tilde O(HS\\sqrt{AT})$ bounds on\nexpected regret under a Bayesian setting, where $S$ and $A$ are the sizes of\nthe state and action spaces, $T$ is time, and $H$ is the bound of the span.\nThis regret bound matches the best available bound for weakly communicating\nMDPs. Numerical results show it to perform better than existing algorithms for\ninfinite horizon MDPs.\n",
"title": "Learning Unknown Markov Decision Processes: A Thompson Sampling Approach"
}
| null | null | null | null | true | null |
9109
| null |
Default
| null | null |
null |
{
"abstract": " The study of knots and links from a probabilistic viewpoint provides insight\ninto the behavior of \"typical\" knots, and opens avenues for new constructions\nof knots and other topological objects with interesting properties. The\nknotting of random curves arises also in applications to the natural sciences,\nsuch as in the context of the structure of polymers. We present here several\nknown and new randomized models of knots and links. We review the main known\nresults on the knot distribution in each model. We discuss the nature of these\nmodels and the properties of the knots they produce. Of particular interest to\nus are finite type invariants of random knots, and the recently studied\nPetaluma model. We report on rigorous results and numerical experiments\nconcerning the asymptotic distribution of such knot invariants. Our approach\nraises questions of universality and classification of the various random knot\nmodels.\n",
"title": "Models of Random Knots"
}
| null | null | null | null | true | null |
9110
| null |
Default
| null | null |
null |
{
"abstract": " The revival structures for the X_m exceptional orthogonal polynomials of the\nScarf I potential endowed with position-dependent effective mass is studied in\nthe context of the generalized Gazeau-Klauder coherent states. It is shown that\nin the case of the constant mass, the deduced coherent states mimic full and\nfractional revivals phenomena. However in the case of position-dependent\neffective mass, although full revivals take place during their time evolution,\nthere is no fractional revivals as defined in the common sense. These\nproperties are illustrated numerically by means of some specific profile mass\nfunctions, with and without singularities. We have also observed a close\nconnection between the coherence time {\\tau}_coh^m? and the mass parameter ?.\n",
"title": "Revival structures of coherent states for Xm exceptional orthogonal polynomials of the Scarf I potential within position-dependent effective mass"
}
| null | null | null | null | true | null |
9111
| null |
Default
| null | null |
null |
{
"abstract": " Autonomous aerial cinematography has the potential to enable automatic\ncapture of aesthetically pleasing videos without requiring human intervention,\nempowering individuals with the capability of high-end film studios. Current\napproaches either only handle off-line trajectory generation, or offer\nstrategies that reason over short time horizons and simplistic representations\nfor obstacles, which result in jerky movement and low real-life applicability.\nIn this work we develop a method for aerial filming that is able to trade off\nshot smoothness, occlusion, and cinematography guidelines in a principled\nmanner, even under noisy actor predictions. We present a novel algorithm for\nreal-time covariant gradient descent that we use to efficiently find the\ndesired trajectories by optimizing a set of cost functions. Experimental\nresults show that our approach creates attractive shots, avoiding obstacles and\nocclusion 65 times over 1.25 hours of flight time, re-planning at 5 Hz with a\n10 s time horizon. We robustly film human actors, cars and bicycles performing\ndifferent motion among obstacles, using various shot types.\n",
"title": "Autonomous drone cinematographer: Using artistic principles to create smooth, safe, occlusion-free trajectories for aerial filming"
}
| null | null | null | null | true | null |
9112
| null |
Default
| null | null |
null |
{
"abstract": " High order reconstruction in the finite volume (FV) approach is achieved by a\nmore fundamental form of the fifth order WENO reconstruction in the framework\nof orthogonally-curvilinear coordinates, for solving the hyperbolic\nconservation equations. The derivation employs a piecewise parabolic polynomial\napproximation to the zone averaged values to reconstruct the right, middle, and\nleft interface values. The grid dependent linear weights of the WENO are\nrecovered by inverting a Vandermode-like linear system of equations with\nspatially varying coefficients. A scheme for calculating the linear weights,\noptimal weights, and smoothness indicator on a regularly- and\nirregularly-spaced grid in orthogonally-curvilinear coordinates is proposed. A\ngrid independent relation for evaluating the smoothness indicator is derived\nfrom the basic definition. Finally, the procedures for the source term\nintegration and extension to multi-dimensions are proposed. Analytical values\nof the linear and optimal weights, and also the weights required for the source\nterm integration and flux averaging, are provided for a regularly-spaced grid\nin Cartesian, cylindrical, and spherical coordinates. Conventional fifth order\nWENO reconstruction for the regularly-spaced grids in the Cartesian coordinates\ncan be fully recovered in the case of limiting curvature. The fifth order\nfinite volume WENO-C (orthogonally-curvilinear version of WENO) reconstruction\nscheme is tested for several 1D and 2D benchmark test cases involving smooth\nand discontinuous flows in cylindrical and spherical coordinates.\n",
"title": "Fifth order finite volume WENO in general orthogonally-curvilinear coordinates"
}
| null | null | null | null | true | null |
9113
| null |
Default
| null | null |
null |
{
"abstract": " Given a statistical model for the request frequencies and sizes of data\nobjects in a caching system, we derive the probability density of the size of\nthe file that accounts for the largest amount of data traffic. This is\nequivalent to finding the required size of the cache for a caching placement\nthat maximizes the expected byte hit ratio for given file size and popularity\ndistributions. The file that maximizes the expected byte hit ratio is the file\nfor which the product of its size and popularity is the highest -- thus, it is\nthe file that incurs the greatest load on the network. We generalize this\ntheoretical problem to cover factors and addends of arbitrary order statistics\nfor given parent distributions. Further, we study the asymptotic behavior of\nthese distributions. We give several factor and addend densities of widely-used\ndistributions, and verify our results by extensive computer simulations.\n",
"title": "Traffic Minimizing Caching and Latent Variable Distributions of Order Statistics"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9114
| null |
Validated
| null | null |
null |
{
"abstract": " We use the language of uninformative Bayesian prior choice to study the\nselection of appropriately simple effective models. We advocate for the prior\nwhich maximizes the mutual information between parameters and predictions,\nlearning as much as possible from limited data. When many parameters are poorly\nconstrained by the available data, we find that this prior puts weight only on\nboundaries of the parameter manifold. Thus it selects a lower-dimensional\neffective theory in a principled way, ignoring irrelevant parameter directions.\nIn the limit where there is sufficient data to tightly constrain any number of\nparameters, this reduces to Jeffreys prior. But we argue that this limit is\npathological when applied to the hyper-ribbon parameter manifolds generic in\nscience, because it leads to dramatic dependence on effects invisible to\nexperiment.\n",
"title": "Maximizing the information learned from finite data selects a simple model"
}
| null | null | null | null | true | null |
9115
| null |
Default
| null | null |
null |
{
"abstract": " Aging, the process of growing old or maturing, is one of the most widely seen\nnatural phenomena in the world. For the stochastic processes, sometimes the\ninfluence of aging can not be ignored. For example, in this paper, by analyzing\nthe functional distribution of the trajectories of aging particles performing\nanomalous diffusion, we reveal that for the fraction of the occupation time\n$T_+/t$ of strong aging particles, $\\langle (T^+(t)^2)\\rangle=\\frac{1}{2}t^2$\nwith coefficient $\\frac{1}{2}$, having no relation with the aging time $t_a$\nand $\\alpha$ and being completely different from the case of weak (none) aging.\nIn fact, we first build the models governing the corresponding functional\ndistributions, i.e., the aging forward and backward Feynman-Kac equations; the\nabove result is one of the applications of the models. Another application of\nthe models is to solve the asymptotic behaviors of the distribution of the\nfirst passage time, $g(t_a,t)$. The striking discovery is that for weakly aging\nsystems, $g(t_a,t)\\sim t_a^{\\frac{\\alpha}{2}}t^{-1-\\frac{\\alpha}{2}}$, while\nfor strongly aging systems, $g(t_a,t)$ behaves as $ t_a^{\\alpha-1}t^{-\\alpha}$.\n",
"title": "Aging Feynman-Kac Equation"
}
| null | null | null | null | true | null |
9116
| null |
Default
| null | null |
null |
{
"abstract": " We present and apply a general-purpose, multi-start algorithm for improving\nthe performance of low-energy samplers used for solving optimization problems.\nThe algorithm iteratively fixes the value of a large portion of the variables\nto values that have a high probability of being optimal. The resulting problems\nare smaller and less connected, and samplers tend to give better low-energy\nsamples for these problems. The algorithm is trivially parallelizable, since\neach start in the multi-start algorithm is independent, and could be applied to\nany heuristic solver that can be run multiple times to give a sample. We\npresent results for several classes of hard problems solved using simulated\nannealing, path-integral quantum Monte Carlo, parallel tempering with\nisoenergetic cluster moves, and a quantum annealer, and show that the success\nmetrics as well as the scaling are improved substantially. When combined with\nthis algorithm, the quantum annealer's scaling was substantially improved for\nnative Chimera graph problems. In addition, with this algorithm the scaling of\nthe time to solution of the quantum annealer is comparable to the Hamze--de\nFreitas--Selby algorithm on the weak-strong cluster problems introduced by\nBoixo et al. Parallel tempering with isoenergetic cluster moves was able to\nconsistently solve 3D spin glass problems with 8000 variables when combined\nwith our method, whereas without our method it could not solve any.\n",
"title": "Effective optimization using sample persistence: A case study on quantum annealers and various Monte Carlo optimization methods"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
9117
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce a novel method to train agents of reinforcement learning (RL) by\nsharing knowledge in a way similar to the concept of using a book. The recorded\ninformation in the form of a book is the main means by which humans learn\nknowledge. Nevertheless, the conventional deep RL methods have mainly focused\neither on experiential learning where the agent learns through interactions\nwith the environment from the start or on imitation learning that tries to\nmimic the teacher. Contrary to these, our proposed book learning shares key\ninformation among different agents in a book-like manner by delving into the\nfollowing two characteristic features: (1) By defining the linguistic function,\ninput states can be clustered semantically into a relatively small number of\ncore clusters, which are forwarded to other RL agents in a prescribed manner.\n(2) By defining state priorities and the contents for recording, core\nexperiences can be selected and stored in a small container. We call this\ncontainer as `BOOK'. Our method learns hundreds to thousand times faster than\nthe conventional methods by learning only a handful of core cluster\ninformation, which shows that deep RL agents can effectively learn through the\nshared knowledge from other agents.\n",
"title": "BOOK: Storing Algorithm-Invariant Episodes for Deep Reinforcement Learning"
}
| null | null | null | null | true | null |
9118
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we set forth a 3-D ocean model of the radioactive trace\nisotopes Th-230 and Pa-231. The interest arises from the fact that these\nisotopes are extensively used for investigating particle transport in the ocean\nand reconstructing past ocean circulation. The tracers are reversibly scavenged\nby biogenic and lithogenic particles.\nOur simulations of Th-230 and Pa-231 are based on the NEMO-PISCES ocean\nbiogeochemistry general circulation model, which includes biogenic particles,\nnamely small and big particulate organic carbon, calcium carbonate and biogenic\nsilica. Small and big lithogenic particles from dust deposition are included in\nour model as well. Their distributions generally compare well with the small\nand big lithogenic particle concentrations from recent observations from the\nGEOTRACES programme, except for boundary nepheloid layers for which, as up to\ntoday, there are no non-trivial, prognostic models available on a global scale.\nOur simulations reproduce Th-230 and Pa-231 dissolved concentrations: they\ncompare well with recent GEOTRACES observations in many parts of the ocean.\nParticulate Th-230 and Pa-231 concentrations are significantly improved\ncompared to previous studies, but they are still too low because of missing\nparticles from nepheloid layers. Our simulation reproduces the main\ncharacteristics of the Pa-231/Th-230 ratio observed in the sediments, and\nsupports a moderate affinity of Pa-231 to biogenic silica as suggested by\nrecent observations, relative to Th-230.\nFuture model development may further improve understanding, especially when\nthis will include a more complete representation of all particles, including\ndifferent size classes, manganese hydroxides and nepheloid layers. This can be\ndone based on our model, as its source code is readily available.\n",
"title": "A global scavenging and circulation ocean model of thorium-230 and protactinium-231 with realistic particle dynamics (NEMO-ProThorP 0.1)"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
9119
| null |
Validated
| null | null |
null |
{
"abstract": " Girard's Geometry of Interaction (GoI), a semantics designed for linear logic\nproofs, has been also successfully applied to programming language semantics.\nOne way is to use abstract machines that pass a token on a fixed graph along a\npath indicated by the GoI. These token-passing abstract machines are space\nefficient, because they handle duplicated computation by repeating the same\nmoves of a token on the fixed graph. Although they can be adapted to obtain\nsound models with regard to the equational theories of various evaluation\nstrategies for the lambda calculus, it can be at the expense of significant\ntime costs. In this paper we show a token-passing abstract machine that can\nimplement evaluation strategies for the lambda calculus, with certified time\nefficiency. Our abstract machine, called the Dynamic GoI Machine (DGoIM),\nrewrites the graph to avoid replicating computation, using the token to find\nthe redexes. The flexibility of interleaving token transitions and graph\nrewriting allows the DGoIM to balance the trade-off of space and time costs.\nThis paper shows that the DGoIM can implement call-by-need evaluation for the\nlambda calculus by using a strategy of interleaving token passing with as much\ngraph rewriting as possible. Our quantitative analysis confirms that the DGoIM\nwith this strategy of interleaving the two kinds of possible operations on\ngraphs can be classified as \"efficient\" following Accattoli's taxonomy of\nabstract machines.\n",
"title": "The Dynamic Geometry of Interaction Machine: A Call-by-need Graph Rewriter"
}
| null | null | null | null | true | null |
9120
| null |
Default
| null | null |
null |
{
"abstract": " We determine the structure of the W-group $\\mathcal{G}_F$, the small Galois\nquotient of the absolute Galois group $G_F$ of the Pythagorean formally real\nfield $F$ when the space of orderings $X_F$ has finite order. Based on\nMarshall's work (1979), we reduce the structure of $\\mathcal{G}_F$ to that of\n$\\mathcal{G}_{\\bar{F}}$, the W-group of the residue field $\\bar{F}$ when $X_F$\nis a connected space. In the disconnected case, the structure of\n$\\mathcal{G}_F$ is the free product of the W-groups $\\mathcal{G}_{F_i}$\ncorresponding to the connected components $X_i$ of $X_F$. We also give a\ncompletely Galois theoretic proof for Marshall's Basic Lemma.\n",
"title": "Classification of finite W-groups"
}
| null | null | null | null | true | null |
9121
| null |
Default
| null | null |
null |
{
"abstract": " Principal component pursuit (PCP) is a state-of-the-art approach for\nbackground estimation problems. Due to their higher computational cost, PCP\nalgorithms, such as robust principal component analysis (RPCA) and its\nvariants, are not feasible in processing high definition videos. To avoid the\ncurse of dimensionality in those algorithms, several methods have been proposed\nto solve the background estimation problem in an incremental manner. We propose\na batch-incremental background estimation model using a special weighted\nlow-rank approximation of matrices. Through experiments with real and synthetic\nvideo sequences, we demonstrate that our method is superior to the\nstate-of-the-art background estimation algorithms such as GRASTA, ReProCS,\nincPCP, and GFL.\n",
"title": "A Batch-Incremental Video Background Estimation Model using Weighted Low-Rank Approximation of Matrices"
}
| null | null | null | null | true | null |
9122
| null |
Default
| null | null |
null |
{
"abstract": " Kepler photometry of the hot Neptune host star HAT-P-11 suggests that its\nspot latitude distribution is comparable to the Sun's near solar maximum. We\nsearch for evidence of an activity cycle in the CaII H & K chromospheric\nemission $S$-index with archival Keck/HIRES spectra and observations from the\nechelle spectrograph on the ARC 3.5 m Telescope at APO. The chromospheric\nemission of HAT-P-11 is consistent with a $\\gtrsim 10$ year activity cycle,\nwhich plateaued near maximum during the Kepler mission. In the cycle that we\nobserved, the star seemed to spend more time near active maximum than minimum.\nWe compare the $\\log R^\\prime_{HK}$ normalized chromospheric emission index of\nHAT-P-11 with other stars. HAT-P-11 has unusually strong chromospheric emission\ncompared to planet-hosting stars of similar effective temperature and rotation\nperiod, perhaps due to tides raised by its planet.\n",
"title": "Chromospheric Activity of HAT-P-11: an Unusually Active Planet-Hosting K Star"
}
| null | null |
[
"Physics"
] | null | true | null |
9123
| null |
Validated
| null | null |
null |
{
"abstract": " While a number of weak consistency mechanisms have been developed in recent\nyears to improve performance and ensure availability in distributed, replicated\nsystems, ensuring correctness of transactional applications running on top of\nsuch systems remains a difficult and important problem. Serializability is a\nwell-understood correctness criterion for transactional programs; understanding\nwhether applications are serializable when executed in a weakly-consistent\nenvironment, however remains a challenging exercise. In this work, we combine\nthe dependency graph-based characterization of serializability and the\nframework of abstract executions to develop a fully automated approach for\nstatically finding bounded serializability violations under \\emph{any} weak\nconsistency model. We reduce the problem of serializability to satisfiability\nof a formula in First-Order Logic, which allows us to harness the power of\nexisting SMT solvers. We provide rules to automatically construct the FOL\nencoding from programs written in SQL (allowing loops and conditionals) and the\nconsistency specification written as a formula in FOL. In addition to detecting\nbounded serializability violations, we also provide two orthogonal schemes to\nreason about unbounded executions by providing sufficient conditions (in the\nform of FOL formulae) whose satisfiability would imply the absence of anomalies\nin any arbitrary execution. We have applied the proposed technique on TPC-C, a\nreal world database program with complex application logic, and were able to\ndiscover anomalies under Parallel Snapshot Isolation, and verify\nserializability for unbounded executions under Snapshot Isolation, two\nconsistency mechanisms substantially weaker than serializability.\n",
"title": "Automated Detection of Serializability Violations under Weak Consistency"
}
| null | null | null | null | true | null |
9124
| null |
Default
| null | null |
null |
{
"abstract": " The Andreev conductance across 2d normal metal (N)/superconductor (SC)\njunctions with relativistic Dirac spectrum is investigated theoretically in the\nBlonder-Tinkham-Klapwijk formalism. It is shown that for relativistic\nmaterials, due to the Klein tunneling instead of impurity potentials, the local\nstrain in the junction is the key factor that determines the transparency of\nthe junction. The local strain is shown to generate an effective Dirac\n$\\delta$-gauge field. A remarkable suppression of the conductance are observed\nas the strength of the gauge field increases. The behaviors of the conductance\nare in well agreement with the results obtained in the case of 1d N/SC\njunction. We also study the Andreev reflection in a topological material near\nthe chiral-to-helical phase transition in the presence of a local strain. The N\nside of the N/SC junction is modeled by the doped Kane-Mele (KM) model. The SC\nregion is a doped correlated KM t-J (KMtJ) model, which has been shown to\nfeature d+id'-wave spin-singlet pairing. With increasing intrinsic spin-orbit\n(SO) coupling, the doped KMtJ system undergoes a topological phase transition\nfrom the chiral d-wave superconductivity to the spin-Chern superconducting\nphase with helical Majorana fermions at edges. We explore the Andreev\nconductance at the two inequivalent Dirac points, respectively and predict the\ndistinctive behaviors for the Andreev conductance across the topological phase\ntransition. Relevance of our results for the adatom-doped graphene is\ndiscussed.\n",
"title": "Andreev reflection in 2D relativistic materials with realistic tunneling transparency in normal-metal-superconductor junctions"
}
| null | null | null | null | true | null |
9125
| null |
Default
| null | null |
null |
{
"abstract": " Contextual bandits are a form of multi-armed bandit in which the agent has\naccess to predictive side information (known as the context) for each arm at\neach time step, and have been used to model personalized news recommendation,\nad placement, and other applications. In this work, we propose a multi-task\nlearning framework for contextual bandit problems. Like multi-task learning in\nthe batch setting, the goal is to leverage similarities in contexts for\ndifferent arms so as to improve the agent's ability to predict rewards from\ncontexts. We propose an upper confidence bound-based multi-task learning\nalgorithm for contextual bandits, establish a corresponding regret bound, and\ninterpret this bound to quantify the advantages of learning in the presence of\nhigh task (arm) similarity. We also describe an effective scheme for estimating\ntask similarity from data, and demonstrate our algorithm's performance on\nseveral data sets.\n",
"title": "Multi-Task Learning for Contextual Bandits"
}
| null | null | null | null | true | null |
9126
| null |
Default
| null | null |
null |
{
"abstract": " Let $Y$ and $Z$ be two given topological spaces, ${\\cal O}(Y)$ (respectively,\n${\\cal O}(Z)$) the set of all open subsets of $Y$ (respectively, $Z$), and\n$C(Y,Z)$ the set of all continuous maps from $Y$ to $Z$. We study Scott type\ntopologies on ${\\mathcal O}(Y)$ and we construct admissible topologies on\n$C(Y,Z)$ and ${\\mathcal O}_Z(Y)=\\{f^{-1}(U)\\in {\\mathcal O}(Y): f\\in C(Y,Z)\\\n{\\rm and}\\ U\\in {\\mathcal O}(Z)\\}$, introducing new problems in the field.\n",
"title": "Admissible topologies on $C(Y,Z)$ and ${\\cal O}_Z(Y)$"
}
| null | null | null | null | true | null |
9127
| null |
Default
| null | null |
null |
{
"abstract": " We consider the spherical mean generated by a multidimensional generalized\ntranslation and general Euler-Poisson-Darboux equation corresponding to this\nmean. The Asgeirsson property of solutions of the ultrahyperbolic equation that\nincludes singular differential Bessel operators acting by each variable is\nprovided.\n",
"title": "Weighted spherical means generated by generalized translation and general Euler-Poisson-Darboux equation"
}
| null | null | null | null | true | null |
9128
| null |
Default
| null | null |
null |
{
"abstract": " The Tabu Search (TS) metaheuristic has been proposed for K-Means clustering\nas an alternative to Lloyd's algorithm, which for all its ease of\nimplementation and fast runtime, has the major drawback of being trapped at\nlocal optima. While the TS approach can yield superior performance, it involves\na high computational complexity. Moreover, the difficulty in parameter\nselection in the existing TS approach does not make it any more attractive.\nThis paper presents an alternative, low-complexity formulation of the TS\noptimization procedure for K-Means clustering. This approach does not require\nmany parameter settings. We initially constrain the centers to points in the\ndataset. We then aim at evolving these centers using a unique neighborhood\nstructure that makes use of gradient information of the objective function.\nThis results in an efficient exploration of the search space, after which the\nmeans are refined. The proposed scheme is implemented in MATLAB and tested on\nfour real-world datasets, and it achieves a significant improvement over the\nexisting TS approach in terms of the intra cluster sum of squares and\ncomputational time.\n",
"title": "K-Means Clustering using Tabu Search with Quantized Means"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9129
| null |
Validated
| null | null |
null |
{
"abstract": " The synthetic toggle switch, first proposed by Gardner & Collins [1] is a\nMIMO control system that can be controlled by varying the concentrations of two\ninducer molecules, aTc and IPTG, to achieve a desired level of expression of\nthe two genes it comprises. It has been shown [2] that this can be accomplished\nthrough an open-loop external control strategy where the two inputs are\nselected as mutually exclusive periodic pulse waves of appropriate amplitude\nand duty-cycle. In this paper, we use a recently derived average model of the\ngenetic toggle switch subject to these inputs to synthesize new feedback\ncontrol approaches that adjust the inputs duty-cycle in real-time via two\ndifferent possible strategies, a model based hybrid PI-PWM approach and a\nso-called Zero-Average dynamics (ZAD) controller. The controllers are validated\nin-silico via both deterministic and stochastic simulations (SSA) illustrating\nthe advantages and limitations of each strategy\n",
"title": "In-silico Feedback Control of a MIMO Synthetic Toggle Switch via Pulse-Width Modulation"
}
| null | null | null | null | true | null |
9130
| null |
Default
| null | null |
null |
{
"abstract": " We study Segre varieties associated to Levi-flat subsets in complex manifolds\nand apply them to establish local and global results on the integration of\ntangent holomorphic foliations.\n",
"title": "Holomorphic foliations tangent to Levi-flat subsets"
}
| null | null | null | null | true | null |
9131
| null |
Default
| null | null |
null |
{
"abstract": " Event cameras are a paradigm shift in camera technology. Instead of full\nframes, the sensor captures a sparse set of events caused by intensity changes.\nSince only the changes are transferred, those cameras are able to capture quick\nmovements of objects in the scene or of the camera itself. In this work we\npropose a novel method to perform camera tracking of event cameras in a\npanoramic setting with three degrees of freedom. We propose a direct camera\ntracking formulation, similar to state-of-the-art in visual odometry. We show\nthat the minimal information needed for simultaneous tracking and mapping is\nthe spatial position of events, without using the appearance of the imaged\nscene point. We verify the robustness to fast camera movements and dynamic\nobjects in the scene on a recently proposed dataset and self-recorded\nsequences.\n",
"title": "Real-Time Panoramic Tracking for Event Cameras"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9132
| null |
Validated
| null | null |
null |
{
"abstract": " We apply the nonlinear reconstruction method to simulated halo fields. For\nhalo number density $2.77\\times 10^{-2}$ $(h^{-1} {\\rm Mpc})^{-3}$ at $z=0$,\ncorresponding to the SDSS main sample density, we find the scale where the\nnoise saturates the linear signal is improved to $k\\gtrsim0.36\\ h {\\rm\nMpc}^{-1}$, a factor of $2.29$ improvement in scale, or $12$ in number of\nlinear modes. The improvement is less for higher redshift or lower halo\ndensity. We expect this to substantially improve the BAO accuracy of dense, low\nredshift surveys, including the SDSS main sample, 6dFGS and 21cm intensity\nmapping initiatives.\n",
"title": "Halo nonlinear reconstruction"
}
| null | null | null | null | true | null |
9133
| null |
Default
| null | null |
null |
{
"abstract": " Establishing metallic hydrogen is a goal of intensive theoretical and\nexperimental work since 1935 when Wigner and Hungtinton [1] predicted that\ninsulating molecular hydrogen will dissociate at high pressures and transform\nto a metal. This metal is predicted to be a superconductor with very high\ncritical temperature [2]. In another scenario, the metallization can be\nrealized through overlapping of electronic bands in molecular hydrogen in the\nsimilar 400 - 500 GPa pressure range [3-5]. The calculations are not accurate\nenough to predict which option will be realized. Our data are consistent with\ntransforms of hydrogen to semimetal by closing the indirect band gap in the\nmolecular phase III at pressure ~ 360 GPa. Above this pressure, the metallic\nbehaviour in the electrical conductivity appears, the reflection significantly\nincreases. With pressure, the electrical conductivity strongly increases as\nmeasured up to 440 GPa. The Raman measurements evidence that hydrogen is in the\nmolecular phase III at pressures at least up to 440 GPa. At higher pressures\nmeasured up to 480 GPa, the Raman signal gradually disappears indicating\nfurther transformation to a good molecular metal or to an atomic state.\n",
"title": "Molecular semimetallic hydrogen"
}
| null | null | null | null | true | null |
9134
| null |
Default
| null | null |
null |
{
"abstract": " The relative performance of competing point forecasts is usually measured in\nterms of loss or scoring functions. It is widely accepted that these scoring\nfunction should be strictly consistent in the sense that the expected score is\nminimized by the correctly specified forecast for a certain statistical\nfunctional such as the mean, median, or a certain risk measure. Thus, strict\nconsistency opens the way to meaningful forecast comparison, but is also\nimportant in regression and M-estimation. Usually strictly consistent scoring\nfunctions for an elicitable functional are not unique. To give guidance on the\nchoice of a scoring function, this paper introduces two additional quality\ncriteria. Order-sensitivity opens the possibility to compare two deliberately\nmisspecified forecasts given that the forecasts are ordered in a certain sense.\nOn the other hand, equivariant scoring functions obey similar equivariance\nproperties as the functional at hand - such as translation invariance or\npositive homogeneity. In our study, we consider scoring functions for popular\nfunctionals, putting special emphasis on vector-valued functionals, e.g. the\npair (mean, variance) or (Value at Risk, Expected Shortfall).\n",
"title": "Order-Sensitivity and Equivariance of Scoring Functions"
}
| null | null | null | null | true | null |
9135
| null |
Default
| null | null |
null |
{
"abstract": " We present in this paper a generic and parameter-free algorithm to\nefficiently build a wide variety of optical components, such as mirrors or\nlenses, that satisfy some light energy constraints. In all of our problems, one\nis given a collimated or point light source and a desired illumination after\nreflection or refraction and the goal is to design the geometry of a mirror or\nlens which transports exactly the light emitted by the source onto the target.\nWe first propose a general framework and show that eight different optical\ncomponent design problems amount to solving a light energy conservation\nequation that involves the computation of visibility diagrams. We then show\nthat these diagrams all have the same structure and can be obtained by\nintersecting a 3D Power diagram with a planar or spherical domain. This allows\nus to propose an efficient and fully generic algorithm capable to solve these\neight optical component design problems. The support of the prescribed target\nillumination can be a set of directions or a set of points located at a finite\ndistance. Our solutions satisfy design constraints such as convexity or\nconcavity. We show the effectiveness of our algorithm on simulated and\nfabricated examples.\n",
"title": "Light in Power: A General and Parameter-free Algorithm for Caustic Design"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9136
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, we investigate the potential of estimating the soil-moisture\ncontent based on VNIR hyperspectral data combined with LWIR data. Measurements\nfrom a multi-sensor field campaign represent the benchmark dataset which\ncontains measured hyperspectral, LWIR, and soil-moisture data conducted on\ngrassland site. We introduce a regression framework with three steps consisting\nof feature selection, preprocessing, and well-chosen regression models. The\nlatter are mainly supervised machine learning models. An exception are the\nself-organizing maps which combine unsupervised and supervised learning. We\nanalyze the impact of the distinct preprocessing methods on the regression\nresults. Of all regression models, the extremely randomized trees model without\npreprocessing provides the best estimation performance. Our results reveal the\npotential of the respective regression framework combined with the VNIR\nhyperspectral data to estimate soil moisture measured under real-world\nconditions. In conclusion, the results of this paper provide a basis for\nfurther improvements in different research directions.\n",
"title": "Developing a machine learning framework for estimating soil moisture with VNIR hyperspectral data"
}
| null | null | null | null | true | null |
9137
| null |
Default
| null | null |
null |
{
"abstract": " Deep neural networks have proved to be a very effective way to perform\nclassification tasks. They excel when the input data is high dimensional, the\nrelationship between the input and the output is complicated, and the number of\nlabeled training examples is large. But it is hard to explain why a learned\nnetwork makes a particular classification decision on a particular test case.\nThis is due to their reliance on distributed hierarchical representations. If\nwe could take the knowledge acquired by the neural net and express the same\nknowledge in a model that relies on hierarchical decisions instead, explaining\na particular decision would be much easier. We describe a way of using a\ntrained neural net to create a type of soft decision tree that generalizes\nbetter than one learned directly from the training data.\n",
"title": "Distilling a Neural Network Into a Soft Decision Tree"
}
| null | null | null | null | true | null |
9138
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we propose an implicit force control scheme for a one-link\nflexible manipulator that interact with a compliant environment. The controller\nwas based in the mathematical model of the manipulator, considering the\ndynamics of the beam flexible and the gravitational force. With this method,\nthe controller parameters are obtained from the structural parameters of the\nbeam (link) of the manipulator. This controller ensure the stability based in\nthe Lyapunov Theory. The controller proposed has two closed loops: the inner\nloop is a tracking control with gravitational force and vibration frequencies\ncompensation and the outer loop is a implicit force control. To evaluate the\nperformance of the controller, we have considered to three different\nmanipulators (the length, the diameter were modified) and three environments\nwith compliance modified. The results obtained from simulations verify the\nasymptotic tracking and regulated in position and force respectively and the\nvibrations suppression of the beam in a finite time.\n",
"title": "A General Scheme Implicit Force Control for a Flexible-Link Manipulator"
}
| null | null | null | null | true | null |
9139
| null |
Default
| null | null |
null |
{
"abstract": " Casual conversations involving multiple speakers and noises from surrounding\ndevices are part of everyday environments and pose challenges for automatic\nspeech recognition systems. These challenges in speech recognition are target\nfor the CHiME-5 challenge. In the present study, an attempt is made to overcome\nthese challenges by employing a convolutional neural network (CNN)-based\nmultichannel end-to-end speech recognition system. The system comprises an\nattention-based encoder-decoder neural network that directly generates a text\nas an output from a sound input. The mulitchannel CNN encoder, which uses\nresidual connections and batch renormalization, is trained with augmented data,\nincluding white noise injection. The experimental results show that the word\nerror rate (WER) was reduced by 11.9% absolute from the end-to-end baseline.\n",
"title": "CNN-based MultiChannel End-to-End Speech Recognition for everyday home environments"
}
| null | null | null | null | true | null |
9140
| null |
Default
| null | null |
null |
{
"abstract": " Tool manipulation is vital for facilitating robots to complete challenging\ntask goals. It requires reasoning about the desired effect of the task and thus\nproperly grasping and manipulating the tool to achieve the task. Task-agnostic\ngrasping optimizes for grasp robustness while ignoring crucial task-specific\nconstraints. In this paper, we propose the Task-Oriented Grasping Network\n(TOG-Net) to jointly optimize both task-oriented grasping of a tool and the\nmanipulation policy for that tool. The training process of the model is based\non large-scale simulated self-supervision with procedurally generated tool\nobjects. We perform both simulated and real-world experiments on two tool-based\nmanipulation tasks: sweeping and hammering. Our model achieves overall 71.1%\ntask success rate for sweeping and 80.0% task success rate for hammering.\nSupplementary material is available at: bit.ly/task-oriented-grasp\n",
"title": "Learning Task-Oriented Grasping for Tool Manipulation from Simulated Self-Supervision"
}
| null | null | null | null | true | null |
9141
| null |
Default
| null | null |
null |
{
"abstract": " Consider a surface $S$ and let $M\\subset S$. If $S\\setminus M$ is not\nconnected, then we say $M$ \\emph{separates} $S$, and we refer to $M$ as a\n\\emph{separating set} of $S$. If $M$ separates $S$, and no proper subset of $M$\nseparates $S$, then we say $M$ is a \\emph{minimal separating set} of $S$. In\nthis paper we use methods of computational combinatorial topology to classify\nthe minimal separating sets of the orientable surfaces of genus $g=2$ and\n$g=3$. The classification for genus 0 and 1 was done in earlier work, using\nmethods of algebraic topology.\n",
"title": "Classification of Minimal Separating Sets in Low Genus Surfaces"
}
| null | null | null | null | true | null |
9142
| null |
Default
| null | null |
null |
{
"abstract": " Non-invasive steady-state visual evoked potential (SSVEP) based\nbrain-computer interface (BCI) systems offer high bandwidth compared to other\nBCI types and require only minimal calibration and training. Virtual reality\n(VR) has been already validated as effective, safe, affordable and motivating\nfeedback modality for BCI experiments. Augmented reality (AR) enhances the\nphysical world by superimposing informative, context sensitive, computer\ngenerated content. In the context of BCI, AR can be used as a friendlier and\nmore intuitive real-world user interface, thereby facilitating a more seamless\nand goal directed interaction. This can improve practicality and usability of\nBCI systems and may help to compensate for their low bandwidth. In this\nfeasibility study, three healthy participants had to finish a complex\nnavigation task in immersive VR and AR conditions using an online SSVEP BCI.\nTwo out of three subjects were successful in all conditions. To our knowledge,\nthis is the first work to present an SSVEP BCI that operates using target\nstimuli integrated in immersive VR and AR (head-mounted display and camera).\nThis research direction can benefit patients by introducing more intuitive and\neffective real-world interaction (e.g. smart home control). It may also be\nrelevant for user groups that require or benefit from hands free operation\n(e.g. due to temporary situational disability).\n",
"title": "A feasibility study on SSVEP-based interaction with motivating and immersive virtual and augmented reality"
}
| null | null | null | null | true | null |
9143
| null |
Default
| null | null |
null |
{
"abstract": " IUIs aim to incorporate intelligent automated capabilities in human computer\ninteraction, where the net impact is a human-computer interaction that improves\nperformance or usability in critical ways. It also involves designing and\nimplementing an artificial intelligence (AI) component that effectively\nleverages human skills and capabilities, so that human performance with an\napplication excels. IUIs embody capabilities that have traditionally been\nassociated more strongly with humans than with computers: how to perceive,\ninterpret, learn, use language, reason, plan, and decide.\n",
"title": "Intelligent User Interfaces - A Tutorial"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9144
| null |
Validated
| null | null |
null |
{
"abstract": " Graphs are widely used to model execution dependencies in applications. In\nparticular, the NP-complete problem of partitioning a graph under constraints\nreceives enormous attention by researchers because of its applicability in\nmultiprocessor scheduling. We identified the additional constraint of acyclic\ndependencies between blocks when mapping computer vision and imaging\napplications to a heterogeneous embedded multiprocessor. Existing algorithms\nand heuristics do not address this requirement and deliver results that are not\napplicable for our use-case. In this work, we show that this more constrained\nversion of the graph partitioning problem is NP-complete and present heuristics\nthat achieve a close approximation of the optimal solution found by an\nexhaustive search for small problem instances and much better scalability for\nlarger instances. In addition, we can show a positive impact on the schedule of\na real imaging application that improves communication volume and execution\ntime.\n",
"title": "Graph Partitioning with Acyclicity Constraints"
}
| null | null | null | null | true | null |
9145
| null |
Default
| null | null |
null |
{
"abstract": " Most of the efficient sublinear-time indexing algorithms for the\nhigh-dimensional nearest neighbor search problem (NNS) are based on space\npartitions of the ambient space $\\mathbb{R}^d$. Inspired by recent theoretical\nwork on NNS for general metric spaces [Andoni, Naor, Nikolov, Razenshteyn,\nWaingarten STOC 2018, FOCS 2018], we develop a new framework for constructing\nsuch partitions that reduces the problem to balanced graph partitioning\nfollowed by supervised classification. We instantiate this general approach\nwith the KaHIP graph partitioner [Sanders, Schulz SEA 2013] and neural\nnetworks, respectively, to obtain a new partitioning procedure called Neural\nLocality-Sensitive Hashing (Neural LSH). On several standard benchmarks for\nNNS, our experiments show that the partitions found by Neural LSH consistently\noutperform partitions found by quantization- and tree-based methods.\n",
"title": "Learning Sublinear-Time Indexing for Nearest Neighbor Search"
}
| null | null | null | null | true | null |
9146
| null |
Default
| null | null |
null |
{
"abstract": " Recent work has proposed the Lempel-Ziv Jaccard Distance (LZJD) as a method\nto measure the similarity between binary byte sequences for malware\nclassification. We propose and test LZJD's effectiveness as a similarity digest\nhash for digital forensics. To do so we develop a high performance Java\nimplementation with the same command-line arguments as sdhash, making it easy\nto integrate into existing workflows. Our testing shows that LZJD is effective\nfor this task, and significantly outperforms sdhash and ssdeep in its ability\nto match related file fragments and files corrupted with random noise. In\naddition, LZJD is up to 60x faster than sdhash at comparison time.\n",
"title": "Lempel-Ziv Jaccard Distance, an Effective Alternative to Ssdeep and Sdhash"
}
| null | null | null | null | true | null |
9147
| null |
Default
| null | null |
null |
{
"abstract": " We study the category of left unital graded modules over the Steinberg\nalgebra of a graded ample Hausdorff groupoid. In the first part of the paper,\nwe show that this category is isomorphic to the category of unital left modules\nover the Steinberg algebra of the skew-product groupoid arising from the\ngrading. To do this, we show that the Steinberg algebra of the skew product is\ngraded isomorphic to a natural generalisation of the the Cohen-Montgomery smash\nproduct of the Steinberg algebra of the underlying groupoid with the grading\ngroup. In the second part of the paper, we study the minimal (that is,\nirreducible) representations in the category of graded modules of a Steinberg\nalgebra, and establish a connection between the annihilator ideals of these\nminimal representations, and effectiveness of the groupoid.\nSpecialising our results, we produce a representation of the monoid of graded\nfinitely generated projective modules over a Leavitt path algebra. We deduce\nthat the lattice of order-ideals in the $K_0$-group of the Leavitt path algebra\nis isomorphic to the lattice of graded ideals of the algebra. We also\ninvestigate the graded monoid for Kumjian--Pask algebras of row-finite\n$k$-graphs with no sources. We prove that these algebras are graded von Neumann\nregular rings, and record some structural consequences of this.\n",
"title": "Graded Steinberg algebras and their representations"
}
| null | null | null | null | true | null |
9148
| null |
Default
| null | null |
null |
{
"abstract": " We consider the Cauchy problem for the damped wave equation under the initial\nstate that the sum of an initial position and an initial velocity vanishes.\nWhen the initial position is non-zero, non-negative and compactly supported, we\nstudy the large time behavior of the spatial null, critical, maximum and\nminimum sets of the solution. The spatial null set becomes a smooth\nhyper-surface homeomorphic to a sphere after a large enough time. The spatial\ncritical set has at least three points after a large enough time. The set of\nspatial maximum points escapes from the convex hull of the support of the\ninitial position. The set of spatial minimum points consists of one point after\na large time, and the unique spatial minimum point converges to the centroid of\nthe initial position at time infinity.\n",
"title": "Movement of time-delayed hot spots in Euclidean space for special initial states"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9149
| null |
Validated
| null | null |
null |
{
"abstract": " Time series prediction has been studied in a variety of domains. However, it\nis still challenging to predict future series given historical observations and\npast exogenous data. Existing methods either fail to consider the interactions\namong different components of exogenous variables which may affect the\nprediction accuracy, or cannot model the correlations between exogenous data\nand target data. Besides, the inherent temporal dynamics of exogenous data are\nalso related to the target series prediction, and thus should be considered as\nwell. To address these issues, we propose an end-to-end deep learning model,\ni.e., Hierarchical attention-based Recurrent Highway Network (HRHN), which\nincorporates spatio-temporal feature extraction of exogenous variables and\ntemporal dynamics modeling of target variables into a single framework.\nMoreover, by introducing the hierarchical attention mechanism, HRHN can\nadaptively select the relevant exogenous features in different semantic levels.\nWe carry out comprehensive empirical evaluations with various methods over\nseveral datasets, and show that HRHN outperforms the state of the arts in time\nseries prediction, especially in capturing sudden changes and sudden\noscillations of time series.\n",
"title": "Hierarchical Attention-Based Recurrent Highway Networks for Time Series Prediction"
}
| null | null |
[
"Statistics"
] | null | true | null |
9150
| null |
Validated
| null | null |
null |
{
"abstract": " We extend our previous results characterizing the loading properties of a\ndiffusing passive scalar advected by a laminar shear flow in ducts and channels\nto more general cross-sectional shapes, including regular polygons and smoothed\ncorner ducts originating from deformations of ellipses. For the case of the\ntriangle, short time skewness is calculated exactly to be positive, while\nlong-time asymptotics shows it to be negative. Monte-Carlo simulations confirm\nthese predictions, and document the time scale for sign change. Interestingly,\nthe equilateral triangle is the only regular polygon with this property, all\nother polygons possess positive skewness at all times, although this cannot\ncannot be proved on finite times due to the lack of closed form flow solutions\nfor such geometries. Alternatively, closed form flow solutions can be\nconstructed for smooth deformations of ellipses, and illustrate how the\npossibility of multiple sign switching in time is unrelated to domain corners.\nExact conditions relating the median and the skewness to the mean are developed\nwhich guarantee when the sign for the skewness implies front (back) loading\nproperties of the evolving tracer distribution along the pipe. Short and long\ntime asymptotics confirm this condition, and Monte-Carlo simulations verify\nthis at all times.\n",
"title": "Mass distribution and skewness for passive scalar transport in pipes with polygonal and smooth cross-sections"
}
| null | null | null | null | true | null |
9151
| null |
Default
| null | null |
null |
{
"abstract": " For commercial one-sun solar modules, up to 80% of the incoming sunlight may\nbe dissipated as heat, potentially raising the temperature 20 C - 30 C higher\nthan the ambient. In the long term, extreme self-heating erodes efficiency and\nshortens lifetime, thereby dramatically reducing the total energy output.\nTherefore, it is critically important to develop effective and practical (and\npreferably passive) cooling methods to reduce operating temperature of PV\nmodules. In this paper, we explore two fundamental (but often overlooked)\norigins of PV self-heating, namely, sub-bandgap absorption and imperfect\nthermal radiation. The analysis suggests that we redesign the optical\nproperties of the solar module to eliminate parasitic absorption\n(selective-spectral cooling) and enhance thermal emission (radiative cooling).\nOur Comprehensive opto-electro-thermal simulation shows that the proposed\ntechniques would cool the one-sun and low-concentrated terrestrial solar\nmodules up to 10 C and 20 C, respectively. This self-cooling would\nsubstantially extend the lifetime for solar modules, with The corresponding\nincrease in energy yields and reduced LCOE.\n",
"title": "An Optics-Based Approach to Thermal Management of Photovoltaics: Selective-Spectral and Radiative Cooling"
}
| null | null | null | null | true | null |
9152
| null |
Default
| null | null |
null |
{
"abstract": " We present clustering properties from 579,492 Lyman break galaxies (LBGs) at\nz~4-6 over the 100 deg^2 sky (corresponding to a 1.4 Gpc^3 volume) identified\nin early data of the Hyper Suprime-Cam (HSC) Subaru strategic program survey.\nWe derive angular correlation functions (ACFs) of the HSC LBGs with\nunprecedentedly high statistical accuracies at z~4-6, and compare them with the\nhalo occupation distribution (HOD) models. We clearly identify significant ACF\nexcesses in 10\"<$\\theta$<90\", the transition scale between 1- and 2-halo terms,\nsuggestive of the existence of the non-linear halo bias effect. Combining the\nHOD models and previous clustering measurements of faint LBGs at z~4-7, we\ninvestigate dark-matter halo mass (Mh) of the z~4-7 LBGs and its correlation\nwith various physical properties including the star-formation rate (SFR), the\nstellar-to-halo mass ratio (SHMR), and the dark matter accretion rate (dotMh)\nover a wide-mass range of Mh/M$_\\odot$=4x10^10-4x10^12. We find that the SHMR\nincreases from z~4 to 7 by a factor of ~4 at Mh~1x10^11 M$_\\odot$, while the\nSHMR shows no strong evolution in the similar redshift range at Mh~1x10^12\nM$_\\odot$. Interestingly, we identify a tight relation of SFR/dotMh-Mh showing\nno significant evolution beyond 0.15 dex in this wide-mass range over z~4-7.\nThis weak evolution suggests that the SFR/dotMh-Mh relation is a fundamental\nrelation in high-redshift galaxy formation whose star formation activities are\nregulated by the dark matter mass assembly. Assuming this fundamental relation,\nwe calculate the cosmic SFR densities (SFRDs) over z=0-10 (a.k.a. Madau-Lilly\nplot). The cosmic SFRD evolution based on the fundamental relation agrees with\nthe one obtained by observations, suggesting that the cosmic SFRD increase from\nz~10 to 4-2 (decrease from z~4-2 to 0) is mainly driven by the increase of the\nhalo abundance (the decrease of the accretion rate).\n",
"title": "GOLDRUSH. II. Clustering of Galaxies at $z\\sim 4-6$ Revealed with the Half-Million Dropouts Over the 100 deg$^2$ Area Corresponding to 1 Gpc$^3$"
}
| null | null |
[
"Physics"
] | null | true | null |
9153
| null |
Validated
| null | null |
null |
{
"abstract": " We prove the least-area, unit-volume, tetrahedral tile of Euclidean space,\nwithout the assumption of Gallagher et al. that the tiling uses only\norientation-preserving images of the tile. The winner remains Sommerville's\ntype 4v.\n",
"title": "The Least-Area Tetrahedral Tile of Space"
}
| null | null | null | null | true | null |
9154
| null |
Default
| null | null |
null |
{
"abstract": " Motivated by multi-hop communication in unreliable wireless networks, we\npresent a percolation theory for time-varying networks. We develop a\nrenormalization group theory for a prototypical network on a regular grid,\nwhere individual links switch stochastically between active and inactive\nstates. The question whether a given source node can communicate with a\ndestination node along paths of active links is equivalent to a percolation\nproblem. Our theory maps the temporal existence of multi-hop paths on an\neffective two-state Markov process. We show analytically how this Markov\nprocess converges towards a memory-less Bernoulli process as the hop distance\nbetween source and destination node increases. Our work extends classical\npercolation theory to the dynamic case and elucidates temporal correlations of\nmessage losses. Quantification of temporal correlations has implications for\nthe design of wireless communication and control protocols, e.g. in\ncyber-physical systems such as self-organized swarms of drones or smart traffic\nnetworks.\n",
"title": "Renormalization group theory for percolation in time-varying networks"
}
| null | null | null | null | true | null |
9155
| null |
Default
| null | null |
null |
{
"abstract": " We investigate a few-body mixture of two bosonic components, each consisting\nof two particles confined in a quasi one-dimensional harmonic trap. By means of\nexact diagonalization with a correlated basis approach we obtain the low-energy\nspectrum and eigenstates for the whole range of repulsive intra- and\ninter-component interaction strengths. We analyse the eigenvalues as a function\nof the inter-component coupling, covering hereby all the limiting regimes, and\ncharacterize the behaviour in-between these regimes by exploiting the\nsymmetries of the Hamiltonian. Provided with this knowledge we study the\nbreathing dynamics in the linear-response regime by slightly quenching the trap\nfrequency symmetrically for both components. Depending on the choice of\ninteractions strengths, we identify 1 to 3 monopole modes besides the breathing\nmode of the center of mass coordinate. For the uncoupled mixture each monopole\nmode corresponds to the breathing oscillation of a specific relative\ncoordinate. Increasing the inter-component coupling first leads to multi-mode\noscillations in each relative coordinate, which turn into single-mode\noscillations of the same frequency in the composite-fermionization regime.\n",
"title": "Spectral properties and breathing dynamics of a few-body Bose-Bose mixture in a 1D harmonic trap"
}
| null | null | null | null | true | null |
9156
| null |
Default
| null | null |
null |
{
"abstract": " The natural uranium assembly, \"QUINTA\", was irradiated with 2, 4, and 8 GeV\ndeuterons. The $^{232}$Th, $^{127}$I, and $^{129}$I samples have been exposed\nto secondary neutrons produced in the assembly at a 20-cm radial distance from\nthe deuteron beam axis. The spectra of gamma rays emitted by the activated\n$^{232}$Th, $^{127}$I, and $^{129}$I samples have been analyzed and several\ntens of product nuclei have been identified. For each of those products,\nneutron-induced reaction rates have been determined. The transmutation power\nfor the $^{129}$I samples is estimated. Experimental results were compared to\nthose calculated with well-known stochastic and deterministic codes.\n",
"title": "Study of secondary neutron interactions with $^{232}$Th, $^{129}$I, and $^{127}$I nuclei with the uranium assembly \"QUINTA\" at 2, 4, and 8 GeV deuteron beams of the JINR Nuclotron accelerator"
}
| null | null | null | null | true | null |
9157
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\operatorname{Con}(\\mathbf T)\\!\\restriction\\!x$ denote the finite\nconsistency statement \"there are no proofs of contradiction in $\\mathbf T$ with\n$\\leq x$ symbols\". For a large class of natural theories $\\mathbf T$, Pudlák\nhas shown that the lengths of the shortest proofs of\n$\\operatorname{Con}(\\mathbf T)\\!\\restriction\\!n$ in the theory $\\mathbf T$\nitself are bounded by a polynomial in $n$. At the same time he conjectures that\n$\\mathbf T$ does not have polynomial proofs of the finite consistency\nstatements $\\operatorname{Con}(\\mathbf T+\\operatorname{Con}(\\mathbf\nT))\\!\\restriction\\!n$. In contrast we show that Peano arithmetic\n($\\mathbf{PA}$) has polynomial proofs of\n$\\operatorname{Con}(\\mathbf{PA}+\\operatorname{Con}^*(\\mathbf{PA}))\\!\\restriction\\!n$,\nwhere $\\operatorname{Con}^*(\\mathbf{PA})$ is the slow consistency statement for\nPeano arithmetic, introduced by S.-D. Friedman, Rathjen and Weiermann. We also\nobtain a new proof of the result that the usual consistency statement\n$\\operatorname{Con}(\\mathbf{PA})$ is equivalent to $\\varepsilon_0$ iterations\nof slow consistency. Our argument is proof-theoretic, while previous\ninvestigations of slow consistency relied on non-standard models of arithmetic.\n",
"title": "Short Proofs for Slow Consistency"
}
| null | null | null | null | true | null |
9158
| null |
Default
| null | null |
null |
{
"abstract": " In a recent study entitled \"Cell nuclei have lower refractive index and mass\ndensity than cytoplasm\", we provided strong evidence indicating that the\nnuclear refractive index (RI) is lower than the RI of the cytoplasm for several\ncell lines. In a complementary study in 2017, entitled \"Is the nuclear\nrefractive index lower than cytoplasm? Validation of phase measurements and\nimplications for light scattering technologies\", Steelman et al. observed a\nlower nuclear RI also for other cell lines and ruled out methodological error\nsources such as phase wrapping and scattering effects. Recently, Yurkin\ncomposed a comment on these 2 publications, entitled \"How a phase image of a\ncell with nucleus refractive index smaller than that of the cytoplasm should\nlook like?\", putting into question the methods used for measuring the cellular\nand nuclear RI in the aforementioned publications by suggesting that a lower\nnuclear RI would produce a characteristic dip in the measured phase profile in\nsitu. We point out the difficulty of identifying this dip in the presence of\nother cell organelles, noise, or blurring due to the imaging point spread\nfunction. Furthermore, we mitigate Yurkin's concerns regarding the ability of\nthe simple-transmission approximation to compare cellular and nuclear RI by\nanalyzing a set of phase images with a novel, scattering-based approach. We\nconclude that the absence of a characteristic dip in the measured phase\nprofiles does not contradict the usage of the simple-transmission approximation\nfor the determination of the average cellular or nuclear RI. Our response can\nbe regarded as an addition to the response by Steelman, Eldridge and Wax. We\nkindly ask the reader to attend to their thorough ascertainment prior to\nreading our response.\n",
"title": "Response to Comment on \"Cell nuclei have lower refractive index and mass density than cytoplasm\""
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
9159
| null |
Validated
| null | null |
null |
{
"abstract": " The whole enterprise of spin compositions can be recast as simple enumerative\ncombinatoric problems. We show here that enumerative combinatorics\n(EC)\\citep{book:Stanley-2011} is a natural setting for spin composition, and\neasily leads to very general analytic formulae -- many of which hitherto not\npresent in the literature. Based on it, we propose three general methods for\ncomputing spin multiplicities; namely, 1) the multi-restricted composition, 2)\nthe generalized binomial and 3) the generating function methods. Symmetric and\nanti-symmetric compositions of $SU(2)$ spins are also discussed, using\ngenerating functions. Of particular importance is the observation that while\nthe common Clebsch-Gordan decomposition (CGD) -- which considers the spins as\ndistinguishable -- is related to integer compositions, the symmetric and\nanti-symmetric compositions (where one considers the spins as\nindistinguishable) are obtained considering integer partitions. The integers in\nquestion here are none other but the occupation numbers of the\nHolstein-Primakoff bosons.\n\\par The pervasiveness of $q-$analogues in our approach is a testament to the\nfundamental role they play in spin compositions. In the appendix, some new\nresults in the power series representation of Gaussian polynomials (or\n$q-$binomial coefficients) -- relevant to symmetric and antisymmetric\ncompositions -- are presented.\n",
"title": "On the composition of an arbitrary collection of $SU(2)$ spins: An Enumerative Combinatoric Approach"
}
| null | null | null | null | true | null |
9160
| null |
Default
| null | null |
null |
{
"abstract": " Clearly, no one likes webpages with poor quality of experience (QoE). Being\nperceived as slow or fast is a key element in the overall perceived QoE of web\napplications. While extensive effort has been put into optimizing web\napplications (both in industry and academia), not a lot of work exists in\ncharacterizing what aspects of webpage loading process truly influence human\nend-user's perception of the \"Speed\" of a page. In this paper we present\n\"SpeedPerception\", a large-scale web performance crowdsourcing framework\nfocused on understanding the perceived loading performance of above-the-fold\n(ATF) webpage content. Our end goal is to create free open-source benchmarking\ndatasets to advance the systematic analysis of how humans perceive webpage\nloading process. In Phase-1 of our \"SpeedPerception\" study using Internet\nRetailer Top 500 (IR 500) websites\n(this https URL), we found that commonly used\nnavigation metrics such as \"onLoad\" and \"Time To First Byte (TTFB)\" fail (less\nthan 60% match) to represent majority human perception when comparing the speed\nof two webpages. We present a simple 3-variable-based machine learning model\nthat explains the majority end-user choices better (with $87 \\pm 2\\%$\naccuracy). In addition, our results suggest that the time needed by end-users\nto evaluate relative perceived speed of webpage is far less than the time of\nits \"visualComplete\" event.\n",
"title": "Perceived Performance of Webpages In the Wild: Insights from Large-scale Crowdsourcing of Above-the-Fold QoE"
}
| null | null | null | null | true | null |
9161
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the properties of entanglement in one-dimensional fermionic\nlattices at the Fulde-Ferrell-Larkin-Ovchinnikov (FFLO) superfluid regime. By\nanalyzing occupation probabilities, which are concepts closely related to FFLO\nand entanglement, we obtain approximate analytical expressions for the\nspin-flip processes at the FFLO regime. We also apply density matrix\nrenormalization group calculations to obtain the exact ground-state\nentanglement of the system in superfluid and non-superfluid regimes. Our\nresults reveal a breaking pairs avalanche appearing precisely at the\nFFLO-normal phase transition. We find that entanglement is non-monotonic in\nsuperfluid regimes, feature that could be used as a signature of exotic\nsuperfluidity.\n",
"title": "Entanglement and exotic superfluidity in spin-imbalanced lattices"
}
| null | null | null | null | true | null |
9162
| null |
Default
| null | null |
null |
{
"abstract": " We revisit the problem of robust principal component analysis with features\nacting as prior side information. To this aim, a novel, elegant, non-convex\noptimization approach is proposed to decompose a given observation matrix into\na low-rank core and the corresponding sparse residual. Rigorous theoretical\nanalysis of the proposed algorithm results in exact recovery guarantees with\nlow computational complexity. Aptly designed synthetic experiments demonstrate\nthat our method is the first to wholly harness the power of non-convexity over\nconvexity in terms of both recoverability and speed. That is, the proposed\nnon-convex approach is more accurate and faster compared to the best available\nalgorithms for the problem under study. Two real-world applications, namely\nimage classification and face denoising further exemplify the practical\nsuperiority of the proposed method.\n",
"title": "Informed Non-convex Robust Principal Component Analysis with Features"
}
| null | null | null | null | true | null |
9163
| null |
Default
| null | null |
null |
{
"abstract": " We show that a self orbit equivalence of a transitive Anosov flow on a\n$3$-manifold which is homotopic to identity has to either preserve every orbit\nor the Anosov flow is $\\mathbb{R}$-covered and the orbit equivalence has to be\nof a specific type. This result shows that one can remove a relatively\nunnatural assumption in a result of Farrell and Gogolev about the topological\nrigidity of bundles supporting a fiberwise Anosov flow when the fiber is\n$3$-dimensional.\n",
"title": "A note on self orbit equivalences of Anosov flows and bundles with fiberwise Anosov flows"
}
| null | null | null | null | true | null |
9164
| null |
Default
| null | null |
null |
{
"abstract": " We use Chandra X-ray data to measure the metallicity of the intracluster\nmedium (ICM) in 245 massive galaxy clusters selected from X-ray and\nSunyaev-Zel'dovich (SZ) effect surveys, spanning redshifts $0<z<1.2$.\nMetallicities were measured in three different radial ranges, spanning cluster\ncores through their outskirts. We explore trends in these measurements as a\nfunction of cluster redshift, temperature, and surface brightness \"peakiness\"\n(a proxy for gas cooling efficiency in cluster centers). The data at large\nradii (0.5--1 $r_{500}$) are consistent with a constant metallicity, while at\nintermediate radii (0.1-0.5 $r_{500}$) we see a late-time increase in\nenrichment, consistent with the expected production and mixing of metals in\ncluster cores. In cluster centers, there are strong trends of metallicity with\ntemperature and peakiness, reflecting enhanced metal production in the\nlowest-entropy gas. Within the cool-core/sharply peaked cluster population,\nthere is a large intrinsic scatter in central metallicity and no overall\nevolution, indicating significant astrophysical variations in the efficiency of\nenrichment. The central metallicity in clusters with flat surface brightness\nprofiles is lower, with a smaller intrinsic scatter, but increases towards\nlower redshifts. Our results are consistent with other recent measurements of\nICM metallicity as a function of redshift. They reinforce the picture implied\nby observations of uniform metal distributions in the outskirts of nearby\nclusters, in which most of the enrichment of the ICM takes place before cluster\nformation, with significant later enrichment taking place only in cluster\ncenters, as the stellar populations of the central galaxies evolve.\n",
"title": "The Metallicity of the Intracluster Medium Over Cosmic Time: Further Evidence for Early Enrichment"
}
| null | null |
[
"Physics"
] | null | true | null |
9165
| null |
Validated
| null | null |
null |
{
"abstract": " In this work, we consider an extension of graphical models to random graphs,\ntrees, and other objects. To do this, many fundamental concepts for\nmultivariate random variables (e.g., marginal variables, Gibbs distribution,\nMarkov properties) must be extended to other mathematical objects; it turns out\nthat this extension is possible, as we will discuss, if we have a consistent,\ncomplete system of projections on a given object. Each projection defines a\nmarginal random variable, allowing one to specify independence assumptions\nbetween them. Furthermore, these independencies can be specified in terms of a\nsmall subset of these marginal variables (which we call the atomic variables),\nallowing the compact representation of independencies by a directed graph.\nProjections also define factors, functions on the projected object space, and\nhence a projection family defines a set of possible factorizations for a\ndistribution; these can be compactly represented by an undirected graph.\nThe invariances used in graphical models are essential for learning\ndistributions, not just on multivariate random variables, but also on other\nobjects. When they are applied to random graphs and random trees, the result is\na general class of models that is applicable to a broad range of problems,\nincluding those in which the graphs and trees have complicated edge structures.\nThese models need not be conditioned on a fixed number of vertices, as is often\nthe case in the literature for random graphs, and can be used for problems in\nwhich attributes are associated with vertices and edges. For graphs,\napplications include the modeling of molecules, neural networks, and relational\nreal-world scenes; for trees, applications include the modeling of infectious\ndiseases, cell fusion, the structure of language, and the structure of objects\nin visual scenes. Many classic models are particular instances of this\nframework.\n",
"title": "Graphical Models: An Extension to Random Graphs, Trees, and Other Objects"
}
| null | null | null | null | true | null |
9166
| null |
Default
| null | null |
null |
{
"abstract": " In the framework of the application of the Boundary Control method to solving\nthe inverse dynamical problems for the one-dimensional Schrödinger and Dirac\noperators on the half-line and semi-infinite discrete Schrödinger operator,\nwe establish the connections with the method of De Branges: for each of the\nsystem we construct the De Branges space and give a natural dynamical\ninterpretation of all its ingredients: the set of function the De Brange space\nconsists of, the scalar product, the reproducing kernel.\n",
"title": "Boundary Control method and De Branges spaces. Schrödinger equation, Dirac system and Discrete Schrödinger operator"
}
| null | null | null | null | true | null |
9167
| null |
Default
| null | null |
null |
{
"abstract": " The traditional view of the morphology-spin connection is being challenged by\nrecent integral-field-unit observations, as the majority of early-type galaxies\nare found to have a rotational component that is often as large as a dispersion\ncomponent. Mergers are often suspected to be critical in galaxy spin evolution,\nyet the details of their roles are still unclear. We present the first results\non the spin evolution of galaxies in cluster environments through a\ncosmological hydrodynamic simulation. Galaxies spin down globally with cosmic\nevolution. Major (mass ratios > 1/4) and minor (1/4 $\\geq$ mass ratios > 1/50)\nmergers are important contributors to the spin down in particular in massive\ngalaxies. Minor mergers appear to have stronger cumulative effects than major\nmergers. Surprisingly, the dominant driver of galaxy spin down seems to be\nenvironmental effects rather than mergers. However, since multiple processes\nact in combination, it is difficult to separate their individual roles. We\nbriefly discuss the caveats and future studies that are called for.\n",
"title": "On the evolution of galaxy spin in a cosmological hydrodynamic simulation of galaxy clusters"
}
| null | null | null | null | true | null |
9168
| null |
Default
| null | null |
null |
{
"abstract": " We present a method to systematically study multi-photon transmission in one\ndimensional systems comprised of correlated quantum emitters coupled to input\nand output waveguides. Within the Green's function approach of the scattering\nmatrix (S-matrix), we develop a diagrammatic technique to analytically obtain\nthe system's scattering amplitudes while at the same time visualise all the\npossible absorption and emission processes. Our method helps to reduce the\nsignificant effort in finding the general response of a many-body bosonic\nsystem, particularly the nonlinear response embedded in the Green's functions.\nWe demonstrate our proposal through physically relevant examples involving\nscattering of multi-photon states from two-level emitters as well as from\narrays of correlated Kerr nonlinear resonators in the Bose-Hubbard model.\n",
"title": "Diagrammatic Approach to Multiphoton Scattering"
}
| null | null |
[
"Physics"
] | null | true | null |
9169
| null |
Validated
| null | null |
null |
{
"abstract": " A method is presented for solving the discrete-time finite-horizon Linear\nQuadratic Regulator (LQR) problem subject to auxiliary linear equality\nconstraints, such as fixed end-point constraints. The method explicitly\ndetermines an affine relationship between the control and state variables, as\nin standard Riccati recursion, giving rise to feedback control policies that\naccount for constraints. Since the linearly-constrained LQR problem arises\ncommonly in robotic trajectory optimization, having a method that can\nefficiently compute these solutions is important. We demonstrate some of the\nuseful properties and interpretations of said control policies, and we compare\nthe computation time of our method against existing methods.\n",
"title": "Efficient Computation of Feedback Control for Constrained Systems"
}
| null | null | null | null | true | null |
9170
| null |
Default
| null | null |
null |
{
"abstract": " As the size of modern data sets exceeds the disk and memory capacities of a\nsingle computer, machine learning practitioners have resorted to parallel and\ndistributed computing. Given that optimization is one of the pillars of machine\nlearning and predictive modeling, distributed optimization methods have\nrecently garnered ample attention, in particular when either observations or\nfeatures are distributed, but not both. We propose a general stochastic\nalgorithm where observations, features, and gradient components can be sampled\nin a double distributed setting, i.e., with both features and observations\ndistributed. Very technical analyses establish convergence properties of the\nalgorithm under different conditions on the learning rate (diminishing to zero\nor constant). Computational experiments in Spark demonstrate a superior\nperformance of our algorithm versus a benchmark in early iterations of the\nalgorithm, which is due to the stochastic components of the algorithm.\n",
"title": "A Stochastic Large-scale Machine Learning Algorithm for Distributed Features and Observations"
}
| null | null | null | null | true | null |
9171
| null |
Default
| null | null |
null |
{
"abstract": " Hybrid inflation, driven by a Fayet-Iliopoulos (FI) D term, is an intriguing\ninflationary model. In its usual formulation, it however suffers from several\nshortcomings. These pertain to the origin of the FI mass scale, the stability\nof scalar fields during inflation, gravitational corrections in supergravity,\nas well as to the latest constraints from the cosmic microwave background. We\ndemonstrate that these issues can be remedied if D-term inflation is realized\nin the context of strongly coupled supersymmetric gauge theories. We suppose\nthat the D term is generated in consequence of dynamical supersymmetry\nbreaking. Moreover, we assume canonical kinetic terms in the Jordan frame as\nwell as an approximate shift symmetry along the inflaton direction. This\nprovides us with a unified picture of D-term inflation and high-scale\nsupersymmetry breaking. The D term may be associated with a gauged U(1)_B-L, so\nthat the end of inflation spontaneously breaks B-L in the visible sector.\n",
"title": "Unified Model of D-Term Inflation"
}
| null | null | null | null | true | null |
9172
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the recovery of a signal from a set of noisy linear\nprojections (measurements), when such projections are unlabeled, that is, the\ncorrespondence between the measurements and the set of projection vectors\n(i.e., the rows of the measurement matrix) is not known a priori. We consider a\nspecial case of unlabeled sensing referred to as Unlabeled Ordered Sampling\n(UOS) where the ordering of the measurements is preserved. We identify a\nnatural duality between this problem and classical Compressed Sensing (CS),\nwhere we show that the unknown support (location of nonzero elements) of a\nsparse signal in CS corresponds to the unknown indices of the measurements in\nUOS. While in CS it is possible to recover a sparse signal from an\nunder-determined set of linear equations (less equations than the signal\ndimension), successful recovery in UOS requires taking more samples than the\ndimension of the signal. Motivated by this duality, we develop a Restricted\nIsometry Property (RIP) similar to that in CS. We also design a low-complexity\nAlternating Minimization algorithm that achieves a stable signal recovery under\nthe established RIP. We analyze our proposed algorithm for different signal\ndimensions and number of measurements theoretically and investigate its\nperformance empirically via simulations. The results are reminiscent of\nphase-transition similar to that occurring in CS.\n",
"title": "Signal Recovery from Unlabeled Samples"
}
| null | null | null | null | true | null |
9173
| null |
Default
| null | null |
null |
{
"abstract": " We study piecewise linear co-dimension two embeddings of closed oriented\nmanifolds in Euclidean space, and show that any such embedding can always be\nisotoped to be a closed braid as long as the ambient dimension is at most five,\nextending results of Alexander (in ambient dimension three), and Viro and\nindependently Kamada (in ambient dimension four). We also show an analogous\nresult for higher co-dimension embeddings.\n",
"title": "Piecewise linear generalized Alexander's theorem in dimension at most 5"
}
| null | null | null | null | true | null |
9174
| null |
Default
| null | null |
null |
{
"abstract": " The problem of population recovery refers to estimating a distribution based\non incomplete or corrupted samples. Consider a random poll of sample size $n$\nconducted on a population of individuals, where each pollee is asked to answer\n$d$ binary questions. We consider one of the two polling impediments: (a) in\nlossy population recovery, a pollee may skip each question with probability\n$\\epsilon$, (b) in noisy population recovery, a pollee may lie on each question\nwith probability $\\epsilon$. Given $n$ lossy or noisy samples, the goal is to\nestimate the probabilities of all $2^d$ binary vectors simultaneously within\naccuracy $\\delta$ with high probability.\nThis paper settles the sample complexity of population recovery. For lossy\nmodel, the optimal sample complexity is\n$\\tilde\\Theta(\\delta^{-2\\max\\{\\frac{\\epsilon}{1-\\epsilon},1\\}})$, improving the\nstate of the art by Moitra and Saks in several ways: a lower bound is\nestablished, the upper bound is improved and the result depends at most on the\nlogarithm of the dimension. Surprisingly, the sample complexity undergoes a\nphase transition from parametric to nonparametric rate when $\\epsilon$ exceeds\n$1/2$. For noisy population recovery, the sharp sample complexity turns out to\nbe more sensitive to dimension and scales as $\\exp(\\Theta(d^{1/3}\n\\log^{2/3}(1/\\delta)))$ except for the trivial cases of $\\epsilon=0,1/2$ or\n$1$.\nFor both models, our estimators simply compute the empirical mean of a\ncertain function, which is found by pre-solving a linear program (LP).\nCuriously, the dual LP can be understood as Le Cam's method for lower-bounding\nthe minimax risk, thus establishing the statistical optimality of the proposed\nestimators. The value of the LP is determined by complex-analytic methods.\n",
"title": "Sample complexity of population recovery"
}
| null | null | null | null | true | null |
9175
| null |
Default
| null | null |
null |
{
"abstract": " We consider the first exit time of a Shiryaev-Roberts diffusion with constant\npositive drift from the interval $[0,A]$ where $A>0$. We show that the moment\ngenerating function (Laplace transform) of a suitably standardized version of\nthe first exit time converges to that of the unit-mean exponential distribution\nas $A\\to+\\infty$. The proof is explicit in that the moment generating function\nof the first exit time is first expressed analytically and in a closed form,\nand then the desired limit as $A\\to+\\infty$ is evaluated directly. The result\nis of importance in the area of quickest change-point detection, and its\ndiscrete-time counterpart has been previously established - although in a\ndifferent manner - by Pollak and Tartakovsky (2009).\n",
"title": "Asymptotic Exponentiality of the First Exit Time of the Shiryaev-Roberts Diffusion with Constant Positive Drift"
}
| null | null | null | null | true | null |
9176
| null |
Default
| null | null |
null |
{
"abstract": " In this article we study the transfer learning model of action advice under a\nbudget. We focus on reinforcement learning teachers providing action advice to\nheterogeneous students playing the game of Pac-Man under a limited advice\nbudget. First, we examine several critical factors affecting advice quality in\nthis setting, such as the average performance of the teacher, its variance and\nthe importance of reward discounting in advising. The experiments show the\nnon-trivial importance of the coefficient of variation (CV) as a statistic for\nchoosing policies that generate advice. The CV statistic relates variance to\nthe corresponding mean. Second, the article studies policy learning for\ndistributing advice under a budget. Whereas most methods in the relevant\nliterature rely on heuristics for advice distribution we formulate the problem\nas a learning one and propose a novel RL algorithm capable of learning when to\nadvise, adapting to the student and the task at hand. Furthermore, we argue\nthat learning to advise under a budget is an instance of a more generic\nlearning problem: Constrained Exploitation Reinforcement Learning.\n",
"title": "Learning to Teach Reinforcement Learning Agents"
}
| null | null | null | null | true | null |
9177
| null |
Default
| null | null |
null |
{
"abstract": " The barocaloric effect is still an incipient scientific topic, but it has\nbeen attracting an increasing attention in the last years due to the promising\nperspectives for its application in alternative cooling devices. Here, we\npresent giant values of barocaloric entropy change and temperature change\ninduced by low pressures in PDMS elastomer around room temperature. Adiabatic\ntemperature changes of 12.0 K and 28.5 K were directly measured for pressure\nchanges of 173 MPa and 390 MPa, respectively, associated with large normalized\ntemperature changes (~70 K GPa-1). From adiabatic temperature change data, we\nobtained entropy change values larger than 140 J kg-1 K-1. We found barocaloric\neffect values that exceed those previously reported for any promising\nbarocaloric materials from direct measurements of temperature change around\nroom temperature. Our results stimulate the study of the barocaloric effect in\nelastomeric polymers and broaden the pathway to use this effect in solid-state\ncooling technologies.\n",
"title": "Giant room-temperature barocaloric effects in PDMS rubber at low pressures"
}
| null | null | null | null | true | null |
9178
| null |
Default
| null | null |
null |
{
"abstract": " We establish $({\\mathfrak{gl}}_M, {\\mathfrak{gl}}_N)$-dualities between\nquantum Gaudin models with irregular singularities. Specifically, for any $M, N\n\\in {\\mathbb Z}_{\\geq 1}$ we consider two Gaudin models: the one associated\nwith the Lie algebra ${\\mathfrak{gl}}_M$ which has a double pole at infinity\nand $N$ poles, counting multiplicities, in the complex plane, and the same\nmodel but with the roles of $M$ and $N$ interchanged. Both models can be\nrealized in terms of Weyl algebras, i.e., free bosons; we establish that, in\nthis realization, the algebras of integrals of motion of the two models\ncoincide. At the classical level we establish two further generalizations of\nthe duality. First, we show that there is also a duality for realizations in\nterms of free fermions. Second, in the bosonic realization we consider the\nclassical cyclotomic Gaudin model associated with the Lie algebra\n${\\mathfrak{gl}}_M$ and its diagram automorphism, with a double pole at\ninfinity and $2N$ poles, counting multiplicities, in the complex plane. We\nprove that it is dual to a non-cyclotomic Gaudin model associated with the Lie\nalgebra ${\\mathfrak{sp}}_{2N}$, with a double pole at infinity and $M$ simple\npoles in the complex plane. In the special case $N=1$ we recover the well-known\nself-duality in the Neumann model.\n",
"title": "$({\\mathfrak{gl}}_M, {\\mathfrak{gl}}_N)$-Dualities in Gaudin Models with Irregular Singularities"
}
| null | null | null | null | true | null |
9179
| null |
Default
| null | null |
null |
{
"abstract": " We propose a method for efficiently coupling the finite element method with\natomistic simulations, while using molecular dynamics or kinetic Monte Carlo\ntechniques. Our method can dynamically build an optimized unstructured mesh\nthat follows the geometry defined by atomistic data. On this mesh, different\nmultiphysics problems can be solved to obtain distributions of physical\nquantities of interest, which can be fed back to the atomistic system. The\nsimulation flow is optimized to maximize computational efficiency while\nmaintaining good accuracy. This is achieved by providing the modules for a)\noptimization of the density of the generated mesh according to requirements of\na specific geometry and b) efficient extension of the finite element domain\nwithout a need to extend the atomistic one. Our method is organized as an\nopen-source C++ code. In the current implementation, an efficient Laplace\nequation solver for calculation of electric field distribution near rough\natomistic surface demonstrates the capability of the suggested approach.\n",
"title": "Dynamic coupling of a finite element solver to large-scale atomistic simulations"
}
| null | null | null | null | true | null |
9180
| null |
Default
| null | null |
null |
{
"abstract": " When the noise affecting time series is colored with unknown statistics, a\ndifficulty for sinusoid detection is to control the true significance level of\nthe test outcome. This paper investigates the possibility of using training\ndata sets of the noise to improve this control. Specifically, we analyze the\nperformances of various detectors {applied to} periodograms standardized using\ntraining data sets. Emphasis is put on sparse detection in the Fourier domain\nand on the limitation posed by the necessarily finite size of the training sets\navailable in practice. We study the resulting false alarm and detection rates\nand show that standardization leads in some cases to powerful constant false\nalarm rate tests. The study is both analytical and numerical. Although\nanalytical results are derived in an asymptotic regime, numerical results show\nthat theory accurately describes the tests' behaviour for moderately large\nsample sizes. Throughout the paper, an application of the considered\nperiodogram standardization is presented for exoplanet detection in radial\nvelocity data.\n",
"title": "A study of periodograms standardized using training data sets and application to exoplanet detection"
}
| null | null | null | null | true | null |
9181
| null |
Default
| null | null |
null |
{
"abstract": " We present an efficient deep learning technique for the model reduction of\nthe Navier-Stokes equations for unsteady flow problems. The proposed technique\nrelies on the Convolutional Neural Network (CNN) and the stochastic gradient\ndescent method. Of particular interest is to predict the unsteady fluid forces\nfor different bluff body shapes at low Reynolds number. The discrete\nconvolution process with a nonlinear rectification is employed to approximate\nthe mapping between the bluff-body shape and the fluid forces. The deep neural\nnetwork is fed by the Euclidean distance function as the input and the target\ndata generated by the full-order Navier-Stokes computations for primitive bluff\nbody shapes. The convolutional networks are iteratively trained using the\nstochastic gradient descent method with the momentum term to predict the fluid\nforce coefficients of different geometries and the results are compared with\nthe full-order computations. We attempt to provide a physical analogy of the\nstochastic gradient method with the momentum term with the simplified form of\nthe incompressible Navier-Stokes momentum equation. We also construct a direct\nrelationship between the CNN-based deep learning and the Mori-Zwanzig formalism\nfor the model reduction of a fluid dynamical system. A systematic convergence\nand sensitivity study is performed to identify the effective dimensions of the\ndeep-learned CNN process such as the convolution kernel size, the number of\nkernels and the convolution layers. Within the error threshold, the prediction\nbased on our deep convolutional network has a speed-up nearly four orders of\nmagnitude compared to the full-order results and consumes an insignificant\nfraction of computational resources. The proposed CNN-based approximation\nprocedure has a profound impact on the parametric design of bluff bodies and\nthe feedback control of separated flows.\n",
"title": "An Efficient Deep Learning Technique for the Navier-Stokes Equations: Application to Unsteady Wake Flow Dynamics"
}
| null | null | null | null | true | null |
9182
| null |
Default
| null | null |
null |
{
"abstract": " This paper describes an efficient algorithm for computing steady\ntwo-dimensional surface gravity wave in irrotational motion. The algorithm\ncomplexity is O(N log N), N being the number of Fourier modes. The algorithm\nallows the arbitrary precision computation of waves in arbitrary depth, i.e.,\nit works efficiently for Stokes, cnoidal and solitary waves, even for quite\nlarge steepnesses. The method is based on conformal mapping, Babenko equation\nrewritten in a suitable way, pseudo-spectral method and Petviashvili's\niterations. The efficiency of the algorithm is illustrated via some relevant\nnumerical examples. The code is open source, so interested readers can easily\ncheck the claims, use and modify the algorithm.\n",
"title": "Accurate fast computation of steady two-dimensional surface gravity waves in arbitrary depth"
}
| null | null | null | null | true | null |
9183
| null |
Default
| null | null |
null |
{
"abstract": " Most of Python and R scientific packages incorporate compiled scientific\nlibraries to speed up the code and reuse legacy libraries. While several\nsemi-automatic solutions exist to wrap these compiled libraries, the process of\nwrapping a large library is cumbersome and time consuming. In this paper, we\nintroduce AutoWIG, a Python package that wraps automatically compiled libraries\ninto high-level languages using LLVM/Clang technologies and the Mako templating\nengine. Our approach is automatic, extensible, and applies to complex C++\nlibraries, composed of thousands of classes or incorporating modern\nmeta-programming constructs.\n",
"title": "AutoWIG: Automatic Generation of Python Bindings for C++ Libraries"
}
| null | null | null | null | true | null |
9184
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the effect of the incommensurate potential on Weyl semimetal,\nwhich is proposed to be realized in ultracold atomic systems trapped in\nthree-dimensional optical lattices. For the system without the Fermi arc, we\nfind that the Weyl points are robust against the incommensurate potential and\nthe system enters into a metallic phase only when the incommensurate potential\nstrength exceeds a critical value. We unveil the trastition by analysing the\nproperties of wave functions and the density of states as a function of the\nincommensurate potential strength. We also study the system with Fermi arcs and\nfind the Fermi arcs are sensitive against the incommensurate potential and can\nbe destoryed by a weak incommensurate potential.\n",
"title": "Fate of Weyl semimetals in the presence of incommensurate potentials"
}
| null | null | null | null | true | null |
9185
| null |
Default
| null | null |
null |
{
"abstract": " Word equations are an important problem on the intersection of formal\nlanguages and algebra. Given two sequences consisting of letters and variables\nwe are to decide whether there is a substitution for the variables that turns\nthis equation into true equality of strings. The computational complexity of\nthis problem remains unknown, with the best lower and upper bounds being NP and\nPSPACE. Recently, a novel technique of recompression was applied to this\nproblem, simplifying the known proofs and lowering the space complexity to\n(nondeterministic) O(n log n). In this paper we show that word equations are in\nnondeterministic linear space, thus the language of satisfiable word equations\nis context-sensitive. We use the known recompression-based algorithm and\nadditionally employ Huffman coding for letters. The proof, however, uses\nanalysis of how the fragments of the equation depend on each other as well as a\nnew strategy for nondeterministic choices of the algorithm, which uses several\nnew ideas to limit the space occupied by the letters.\n",
"title": "Word equations in linear space"
}
| null | null | null | null | true | null |
9186
| null |
Default
| null | null |
null |
{
"abstract": " This paper addresses the problem of minimum cost resilient\nactuation-sensing-communication co-design for regular descriptor systems while\nensuring selective strong structural system's properties. More specifically,\nthe problem consists of determining the minimum cost deployment of actuation\nand sensing technology, as well as communication between the these, such that\ndecentralized control approaches are viable for an arbitrary realization of\nregular descriptor systems satisfying a pre-specified selective structure,\ni.e., some entries can be zero, nonzero, or either zero/nonzero. Towards this\ngoal, we rely on strong structural systems theory and extend it to cope with\nthe selective structure that casts resiliency/robustness properties and\nuncertainty properties of system's model. Upon such framework, we introduce the\nnotion of selective strong structural fixed modes as a characterization of the\nfeasibility of decentralized control laws. Also, we provide necessary and\nsufficient conditions for this property to hold, and show how these conditions\ncan be leveraged to determine the minimum cost resilient placement of\nactuation-sensing-communication technology ensuring feasible solutions. In\nparticular, we study the minimum cost resilient actuation and sensing\nplacement, upon which we construct the solution to our problem. Finally, we\nillustrate the applicability the main results of this paper on an electric\npower grid example.\n",
"title": "Selective Strong Structural Minimum Cost Resilient Co-Design for Regular Descriptor Linear Systems"
}
| null | null | null | null | true | null |
9187
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a new algorithm for reinforcement learning called Maximum\naposteriori Policy Optimisation (MPO) based on coordinate ascent on a relative\nentropy objective. We show that several existing methods can directly be\nrelated to our derivation. We develop two off-policy algorithms and demonstrate\nthat they are competitive with the state-of-the-art in deep reinforcement\nlearning. In particular, for continuous control, our method outperforms\nexisting methods with respect to sample efficiency, premature convergence and\nrobustness to hyperparameter settings while achieving similar or better final\nperformance.\n",
"title": "Maximum a Posteriori Policy Optimisation"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9188
| null |
Validated
| null | null |
null |
{
"abstract": " The geometric approach to optimal transport and information theory has\ntriggered the interpretation of probability densities as an\ninfinite-dimensional Riemannian manifold. The most studied Riemannian\nstructures are Otto's metric, yielding the $L^2$-Wasserstein distance of\noptimal mass transport, and the Fisher--Rao metric, predominant in the theory\nof information geometry. On the space of smooth probability densities, none of\nthese Riemannian metrics are geodesically complete---a property desirable for\nexample in imaging applications. That is, the existence interval for solutions\nto the geodesic flow equations cannot be extended to the whole real line. Here\nwe study a class of Hamilton--Jacobi-like partial differential equations\narising as geodesic flow equations for higher-order Sobolev type metrics on the\nspace of smooth probability densities. We give order conditions for global\nexistence and uniqueness, thereby providing geodesic completeness. The system\nwe study is an interesting example of a flow equation with loss of derivatives,\nwhich is well-posed in the smooth category, yet non-parabolic and fully\nnon-linear. On a more general note, the paper establishes a link between\ngeometric analysis on the space of probability densities and analysis of\nEuler-Arnold equations in topological hydrodynamics.\n",
"title": "On Geodesic Completeness for Riemannian Metrics on Smooth Probability Densities"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9189
| null |
Validated
| null | null |
null |
{
"abstract": " The history of humanhood has included competitive activities of many\ndifferent forms. Sports have offered many benefits beyond that of\nentertainment. At the time of this article, there exists not a competitive\necosystem for cyber security beyond that of conventional capture the flag\ncompetitions, and the like. This paper introduces a competitive framework with\na foundation on computer science, and hacking. This proposed competitive\nlandscape encompasses the ideas underlying information security, software\nengineering, and cyber warfare. We also demonstrate the opportunity to rank,\nscore, & categorize actionable skill levels into tiers of capability.\nPhysiological metrics are analyzed from participants during gameplay. These\nanalyses provide support regarding the intricacies required for competitive\nplay, and analysis of play. We use these intricacies to build a case for an\norganized competitive ecosystem. Using previous player behavior from gameplay,\nwe also demonstrate the generation of an artificial agent purposed with\ngameplay at a competitive level.\n",
"title": "Hacker Combat: A Competitive Sport from Programmatic Dueling & Cyberwarfare"
}
| null | null | null | null | true | null |
9190
| null |
Default
| null | null |
null |
{
"abstract": " Virtual reality simulation is becoming popular as a training platform in\nsurgical education. However, one important aspect of simulation-based surgical\ntraining that has not received much attention is the provision of automated\nreal-time performance feedback to support the learning process. Performance\nfeedback is actionable advice that improves novice behaviour. In simulation,\nautomated feedback is typically extracted from prediction models trained using\ndata mining techniques. Existing techniques suffer from either low\neffectiveness or low efficiency resulting in their inability to be used in\nreal-time. In this paper, we propose a random forest based method that finds a\nbalance between effectiveness and efficiency. Experimental results in a\ntemporal bone surgery simulation show that the proposed method is able to\nextract highly effective feedback at a high level of efficiency.\n",
"title": "Providing Effective Real-time Feedback in Simulation-based Surgical Training"
}
| null | null | null | null | true | null |
9191
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents a novel deep learning architecture to classify structured\nobjects in datasets with a large number of visually similar categories. We\nmodel sequences of images as linear-chain CRFs, and jointly learn the\nparameters from both local-visual features and neighboring classes. The visual\nfeatures are computed by convolutional layers, and the class embeddings are\nlearned by factorizing the CRF pairwise potential matrix. This forms a highly\nnonlinear objective function which is trained by optimizing a local likelihood\napproximation with batch-normalization. This model overcomes the difficulties\nof existing CRF methods to learn the contextual relationships thoroughly when\nthere is a large number of classes and the data is sparse. The performance of\nthe proposed method is illustrated on a huge dataset that contains images of\nretail-store product displays, taken in varying settings and viewpoints, and\nshows significantly improved results compared to linear CRF modeling and\nunnormalized likelihood optimization.\n",
"title": "Large-Scale Classification of Structured Objects using a CRF with Deep Class Embedding"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9192
| null |
Validated
| null | null |
null |
{
"abstract": " The ability to generate natural language sequences from source code snippets\nhas a variety of applications such as code summarization, documentation, and\nretrieval. Sequence-to-sequence (seq2seq) models, adopted from neural machine\ntranslation (NMT), have achieved state-of-the-art performance on these tasks by\ntreating source code as a sequence of tokens. We present ${\\rm {\\scriptsize\nCODE2SEQ}}$: an alternative approach that leverages the syntactic structure of\nprogramming languages to better encode source code. Our model represents a code\nsnippet as the set of compositional paths in its abstract syntax tree (AST) and\nuses attention to select the relevant paths while decoding. We demonstrate the\neffectiveness of our approach for two tasks, two programming languages, and\nfour datasets of up to $16$M examples. Our model significantly outperforms\nprevious models that were specifically designed for programming languages, as\nwell as state-of-the-art NMT models. An interactive online demo of our model is\navailable at this http URL. Our code, data and trained models are\navailable at this http URL.\n",
"title": "code2seq: Generating Sequences from Structured Representations of Code"
}
| null | null | null | null | true | null |
9193
| null |
Default
| null | null |
null |
{
"abstract": " We explore the problem of learning to decompose spatial tasks into segments,\nas exemplified by the problem of a painting robot covering a large object.\nInspired by the ability of classical decision tree algorithms to construct\nstructured partitions of their input spaces, we formulate the problem of\ndecomposing objects into segments as a parsing approach. We make the insight\nthat the derivation of a parse-tree that decomposes the object into segments\nclosely resembles a decision tree constructed by ID3, which can be done when\nthe ground-truth available. We learn to imitate an expert parsing oracle, such\nthat our neural parser can generalize to parse natural images without ground\ntruth. We introduce a novel deterministic policy gradient update, DRAG (i.e.,\nDeteRministically AGgrevate) in the form of a deterministic actor-critic\nvariant of AggreVaTeD, to train our neural parser. From another perspective,\nour approach is a variant of the Deterministic Policy Gradient suitable for the\nimitation learning setting. The deterministic policy representation offered by\ntraining our neural parser with DRAG allows it to outperform state of the art\nimitation and reinforcement learning approaches.\n",
"title": "Learning Neural Parsers with Deterministic Differentiable Imitation Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9194
| null |
Validated
| null | null |
null |
{
"abstract": " We prove that the zero set of a nonnegative plurisubharmonic function that\nsolves $\\det (\\partial \\overline{\\partial} u) \\geq 1$ in $\\mathbb{C}^n$ and is\nin $W^{2, \\frac{n(n-k)}{k}}$ contains no analytic sub-variety of dimension $k$\nor larger. Along the way we prove an analogous result for the real\nMonge-Ampère equation, which is also new. These results are sharp in view of\nwell-known examples of Pogorelov and B{\\l}ocki. As an application, in the real\ncase we extend interior regularity results to the case that $u$ lies in a\ncritical Sobolev space (or more generally, certain Sobolev-Orlicz spaces).\n",
"title": "Dimension of the minimum set for the real and complex Monge-Ampère equations in critical Sobolev spaces"
}
| null | null | null | null | true | null |
9195
| null |
Default
| null | null |
null |
{
"abstract": " Autonomous driving systems are broadly used equipment in the industries and\nin our daily lives, they assist in production, but are majorly used for\nexploration in dangerous or unfamiliar locations. Thus, for a successful\nexploration, navigation plays a significant role. Road detection is an\nessential factor that assists autonomous robots achieved perfect navigation.\nVarious techniques using camera sensors have been proposed by numerous scholars\nwith inspiring results, but their techniques are still vulnerable to these\nenvironmental noises: rain, snow, light intensity and shadow. In addressing\nthese problems, this paper proposed to enhance the road detection system with\nfiltering algorithm to overcome these limitations. Normalized Differences Index\n(NDI) and morphological operation are the filtering algorithms used to address\nthe effect of shadow and guidance and re-guidance image filtering algorithms\nare used to address the effect of rain and/or snow, while dark channel image\nand specular-to-diffuse are the filters used to address light intensity\neffects. The experimental performance of the road detection system with\nfiltering algorithms was tested qualitatively and quantitatively using the\nfollowing evaluation schemes: False Negative Rate (FNR) and False Positive Rate\n(FPR). Comparison results of the road detection system with and without\nfiltering algorithm shows the filtering algorithm's capability to suppress the\neffect of environmental noises because better road/non-road classification is\nachieved by the road detection system. with filtering algorithm. This\nachievement has further improved path planning/region classification for\nautonomous driving system\n",
"title": "Road Detection Technique Using Filters with Application to Autonomous Driving System"
}
| null | null | null | null | true | null |
9196
| null |
Default
| null | null |
null |
{
"abstract": " Achieving relativistic flight to enable extrasolar exploration is one of the\ndreams of humanity and the long term goal of our NASA Starlight program. We\nderive a fully relativistic solution for the motion of a spacecraft propelled\nby radiation pressure from a directed energy system. Depending on the system\nparameters, low mass spacecraft can achieve relativistic speeds; thereby\nenabling interstellar exploration. The diffraction of the directed energy\nsystem plays an important role and limits the maximum speed of the spacecraft.\nWe consider 'photon recycling' as a possible method to achieving higher speeds.\nWe also discuss recent claims that our previous work on this topic is incorrect\nand show that these claims arise from an improper treatment of causality.\n",
"title": "Relativistic Spacecraft Propelled by Directed Energy"
}
| null | null | null | null | true | null |
9197
| null |
Default
| null | null |
null |
{
"abstract": " Estimating cascade size and nodes' influence is a fundamental task in social,\ntechnological, and biological networks. Yet this task is extremely challenging\ndue to the sheer size and the structural heterogeneity of networks. We\ninvestigate a new influence measure, termed outward influence (OI), defined as\nthe (expected) number of nodes that a subset of nodes $S$ will activate,\nexcluding the nodes in S. Thus, OI equals, the de facto standard measure,\ninfluence spread of S minus |S|. OI is not only more informative for nodes with\nsmall influence, but also, critical in designing new effective sampling and\nstatistical estimation methods.\nBased on OI, we propose SIEA/SOIEA, novel methods to estimate influence\nspread/outward influence at scale and with rigorous theoretical guarantees. The\nproposed methods are built on two novel components 1) IICP an important\nsampling method for outward influence, and 2) RSA, a robust mean estimation\nmethod that minimize the number of samples through analyzing variance and range\nof random variables. Compared to the state-of-the art for influence estimation,\nSIEA is $\\Omega(\\log^4 n)$ times faster in theory and up to several orders of\nmagnitude faster in practice. For the first time, influence of nodes in the\nnetworks of billions of edges can be estimated with high accuracy within a few\nminutes. Our comprehensive experiments on real-world networks also give\nevidence against the popular practice of using a fixed number, e.g. 10K or 20K,\nof samples to compute the \"ground truth\" for influence spread.\n",
"title": "Outward Influence and Cascade Size Estimation in Billion-scale Networks"
}
| null | null | null | null | true | null |
9198
| null |
Default
| null | null |
null |
{
"abstract": " High frequency based estimation methods for a semiparametric pure-jump\nsubordinated Brownian motion exposed to a small additive microstructure noise\nare developed building on the two-scales realized variations approach\noriginally developed by Zhang et. al. (2005) for the estimation of the\nintegrated variance of a continuous Ito process. The proposed estimators are\nshown to be robust against the noise and, surprisingly, to attain better rates\nof convergence than their precursors, method of moment estimators, even in the\nabsence of microstructure noise. Our main results give approximate optimal\nvalues for the number K of regular sparse subsamples to be used, which is an\nimportant tune-up parameter of the method. Finally, a data-driven plug-in\nprocedure is devised to implement the proposed estimators with the optimal\nK-value. The developed estimators exhibit superior performance as illustrated\nby Monte Carlo simulations and a real high-frequency data application.\n",
"title": "Estimation of a noisy subordinated Brownian Motion via two-scales power variations"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
9199
| null |
Validated
| null | null |
null |
{
"abstract": " KAGRA is a 3-km cryogenic interferometric gravitational wave telescope\nlocated at an underground site in Japan. In order to achieve its target\nsensitivity, the relative positions of the mirrors of the interferometer must\nbe finely adjusted with attached actuators. We have developed a model to\nsimulate the length control loops of the KAGRA interferometer with realistic\nsuspension responses and various noises for mirror actuation. Using our model,\nwe have designed the actuation parameters to have sufficient force range to\nacquire lock as well as to control all the length degrees of freedom without\nintroducing excess noise.\n",
"title": "Mirror actuation design for the interferometer control of the KAGRA gravitational wave telescope"
}
| null | null | null | null | true | null |
9200
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.