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": " Deep neural networks (DNNs) have transformed several artificial intelligence\nresearch areas including computer vision, speech recognition, and natural\nlanguage processing. However, recent studies demonstrated that DNNs are\nvulnerable to adversarial manipulations at testing time. Specifically, suppose\nwe have a testing example, whose label can be correctly predicted by a DNN\nclassifier. An attacker can add a small carefully crafted noise to the testing\nexample such that the DNN classifier predicts an incorrect label, where the\ncrafted testing example is called adversarial example. Such attacks are called\nevasion attacks. Evasion attacks are one of the biggest challenges for\ndeploying DNNs in safety and security critical applications such as\nself-driving cars. In this work, we develop new methods to defend against\nevasion attacks. Our key observation is that adversarial examples are close to\nthe classification boundary. Therefore, we propose region-based classification\nto be robust to adversarial examples. For a benign/adversarial testing example,\nwe ensemble information in a hypercube centered at the example to predict its\nlabel. In contrast, traditional classifiers are point-based classification,\ni.e., given a testing example, the classifier predicts its label based on the\ntesting example alone. Our evaluation results on MNIST and CIFAR-10 datasets\ndemonstrate that our region-based classification can significantly mitigate\nevasion attacks without sacrificing classification accuracy on benign examples.\nSpecifically, our region-based classification achieves the same classification\naccuracy on testing benign examples as point-based classification, but our\nregion-based classification is significantly more robust than point-based\nclassification to various evasion attacks.\n",
"title": "Mitigating Evasion Attacks to Deep Neural Networks via Region-based Classification"
}
| null | null | null | null | true | null |
17201
| null |
Default
| null | null |
null |
{
"abstract": " Let $m(n)$ denote the maximum size of a family of subsets which does not\ncontain two disjoint sets along with their union. In 1968 Kleitman proved that\n$m(n) = {n\\choose m+1}+\\ldots +{n\\choose 2m+1}$ if $n=3m+1$. Confirming the\nconjecture of Kleitman, we establish the same equality for the cases $n=3m$ and\n$n=3m+2$, and also determine all extremal families. Unlike the case $n=3m+1$,\nthe extremal families are not unique. This is a plausible reason behind the\nrelative difficulty of our proofs. We completely settle the case of several\nfamilies as well.\n",
"title": "Partition-free families of sets"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17202
| null |
Validated
| null | null |
null |
{
"abstract": " Test mass charging caused by cosmic rays will be a significant source of\nacceleration noise for space-based gravitational wave detectors like LISA.\nOperating between December 2015 and July 2017, the technology demonstration\nmission LISA Pathfinder included a bespoke monitor to help characterise the\nrelationship between test mass charging and the local radiation environment.\nThe radiation monitor made in situ measurements of the cosmic ray flux while\nalso providing information about its energy spectrum. We describe the monitor\nand present measurements which show a gradual 40% increase in count rate\ncoinciding with the declining phase of the solar cycle. Modulations of up to\n10% were also observed with periods of 13 and 26 days that are associated with\nco-rotating interaction regions and heliospheric current sheet crossings. These\nvariations in the flux above the monitor detection threshold (approximately 70\nMeV) are shown to be coherent with measurements made by the IREM monitor\non-board the Earth orbiting INTEGRAL spacecraft. Finally we use the measured\ndeposited energy spectra, in combination with a GEANT4 model, to estimate the\ngalactic cosmic ray differential energy spectrum over the course of the\nmission.\n",
"title": "Measuring the Galactic Cosmic Ray Flux with the LISA Pathfinder Radiation Monitor"
}
| null | null | null | null | true | null |
17203
| null |
Default
| null | null |
null |
{
"abstract": " We propose a memory-model-aware static program analysis method for accurately\nanalyzing the behavior of concurrent software running on processors with weak\nconsistency models such as x86-TSO, SPARC-PSO, and SPARC-RMO. At the center of\nour method is a unified framework for deciding the feasibility of inter-thread\ninterferences to avoid propagating spurious data flows during static analysis\nand thus boost the performance of the static analyzer. We formulate the\nchecking of interference feasibility as a set of Datalog rules which are both\nefficiently solvable and general enough to capture a range of hardware-level\nmemory models. Compared to existing techniques, our method can significantly\nreduce the number of bogus alarms as well as unsound proofs. We implemented the\nmethod and evaluated it on a large set of multithreaded C programs. Our\nexperiments showthe method significantly outperforms state-of-the-art\ntechniques in terms of accuracy with only moderate run-time overhead.\n",
"title": "Thread-Modular Static Analysis for Relaxed Memory Models"
}
| null | null | null | null | true | null |
17204
| null |
Default
| null | null |
null |
{
"abstract": " We present an evaluation update (or simply, update) algorithm for a\nfull-featured functional programming language, which synthesizes program\nchanges based on output changes. Intuitively, the update algorithm retraces the\nsteps of the original evaluation, rewriting the program as needed to reconcile\ndifferences between the original and updated output values. Our approach,\nfurthermore, allows expert users to define custom lenses that augment the\nupdate algorithm with more advanced or domain-specific program updates.\nTo demonstrate the utility of evaluation update, we implement the algorithm\nin Sketch-n-Sketch, a novel direct manipulation programming system for\ngenerating HTML documents. In Sketch-n-Sketch, the user writes an ML-style\nfunctional program to generate HTML output. When the user directly manipulates\nthe output using a graphical user interface, the update algorithm reconciles\nthe changes. We evaluate bidirectional evaluation in Sketch-n-Sketch by\nauthoring ten examples comprising approximately 1400 lines of code in total.\nThese examples demonstrate how a variety of HTML documents and applications can\nbe developed and edited interactively in Sketch-n-Sketch, mitigating the\ntedious edit-run-view cycle in traditional programming environments.\n",
"title": "Bidirectional Evaluation with Direct Manipulation"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17205
| null |
Validated
| null | null |
null |
{
"abstract": " We analyse extreme event statistics of experimentally realized Markov chains\nwith various drifts. Our Markov chains are individual trajectories of a single\natom diffusing in a one dimensional periodic potential. Based on more than 500\nindividual atomic traces we verify the applicability of the Sparre Andersen\ntheorem to our system despite the presence of a drift. We present detailed\nanalysis of four different rare event statistics for our system: the\ndistributions of extreme values, of record values, of extreme value occurrence\nin the chain, and of the number of records in the chain. We observe that for\nour data the shape of the extreme event distributions is dominated by the\nunderlying exponential distance distribution extracted from the atomic traces.\nFurthermore, we find that even small drifts influence the statistics of extreme\nevents and record values, which is supported by numerical simulations, and we\nidentify cases in which the drift can be determined without information about\nthe underlying random variable distributions. Our results facilitate the use of\nextreme event statistics as a signal for small drifts in correlated\ntrajectories.\n",
"title": "Extreme Event Statistics in a Drifting Markov Chain"
}
| null | null | null | null | true | null |
17206
| null |
Default
| null | null |
null |
{
"abstract": " The self-consistent nonlinear interaction of a monoenergetic bunch with cold\nplasma is considered. It is shown that under certain conditions a\nself-acceleration of the bunch tail electrons up to high energies is possible.\n",
"title": "On a Possibility of Self Acceleration of Electrons in a Plasma"
}
| null | null | null | null | true | null |
17207
| null |
Default
| null | null |
null |
{
"abstract": " We present the first adaptive strategy for active learning in the setting of\nclassification with smooth decision boundary. The problem of adaptivity (to\nunknown distributional parameters) has remained opened since the seminal work\nof Castro and Nowak (2007), which first established (active learning) rates for\nthis setting. While some recent advances on this problem establish adaptive\nrates in the case of univariate data, adaptivity in the more practical setting\nof multivariate data has so far remained elusive. Combining insights from\nvarious recent works, we show that, for the multivariate case, a careful\nreduction to univariate-adaptive strategies yield near-optimal rates without\nprior knowledge of distributional parameters.\n",
"title": "An Adaptive Strategy for Active Learning with Smooth Decision Boundary"
}
| null | null | null | null | true | null |
17208
| null |
Default
| null | null |
null |
{
"abstract": " Synthesizing images of the eye fundus is a challenging task that has been\npreviously approached by formulating complex models of the anatomy of the eye.\nNew images can then be generated by sampling a suitable parameter space. In\nthis work, we propose a method that learns to synthesize eye fundus images\ndirectly from data. For that, we pair true eye fundus images with their\nrespective vessel trees, by means of a vessel segmentation technique. These\npairs are then used to learn a mapping from a binary vessel tree to a new\nretinal image. For this purpose, we use a recent image-to-image translation\ntechnique, based on the idea of adversarial learning. Experimental results show\nthat the original and the generated images are visually different in terms of\ntheir global appearance, in spite of sharing the same vessel tree.\nAdditionally, a quantitative quality analysis of the synthetic retinal images\nconfirms that the produced images retain a high proportion of the true image\nset quality.\n",
"title": "Towards Adversarial Retinal Image Synthesis"
}
| null | null | null | null | true | null |
17209
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the growth of the graphene buffer layer and the involved step\nbunching behavior of the silicon carbide substrate surface using atomic force\nmicroscopy. The formation of local buffer layer domains are identified to be\nthe origin of undesirably high step edges in excellent agreement with the\npredictions of a general model of step dynamics. The applied polymer-assisted\nsublimation growth method demonstrates that the key principle to suppress this\nbehavior is the uniform nucleation of the buffer layer. In this way, the\nsilicon carbide surface is stabilized such that ultra-flat surfaces can be\nconserved during graphene growth on a large variety of silicon carbide\nsubstrate surfaces. The analysis of the experimental results describes\ndifferent growth modes which extend the current understanding of epitaxial\ngraphene growth by emphasizing the importance of buffer layer nucleation and\ncritical mass transport processes.\n",
"title": "Tailoring the SiC surface - a morphology study on the epitaxial growth of graphene and its buffer layer"
}
| null | null | null | null | true | null |
17210
| null |
Default
| null | null |
null |
{
"abstract": " Vaccine hesitancy has been recognized as a major global health threat. Having\naccess to any type of information in social media has been suggested as a\npotential powerful influence factor to hesitancy. Recent studies in other\nfields than vaccination show that access to a wide amount of content through\nthe Internet without intermediaries resolved into major segregation of the\nusers in polarized groups. Users select the information adhering to theirs\nsystem of beliefs and tend to ignore dissenting information. In this paper we\nassess whether there is polarization in Social Media use in the field of\nvaccination. We perform a thorough quantitative analysis on Facebook analyzing\n2.6M users interacting with 298.018 posts over a time span of seven years and 5\nmonths. We used community detection algorithms to automatically detect the\nemergent communities from the users activity and to quantify the cohesiveness\nover time of the communities. Our findings show that content consumption about\nvaccines is dominated by the echo-chamber effect and that polarization\nincreased over years. Communities emerge from the users consumption habits,\ni.e. the majority of users only consumes information in favor or against\nvaccines, not both. The existence of echo-chambers may explain why social-media\ncampaigns providing accurate information may have limited reach, may be\neffective only in sub-groups and might even foment further polarization of\nopinions. The introduction of dissenting information into a sub-group is\ndisregarded and can have a backfire effect, further reinforcing the existing\nopinions within the sub-group.\n",
"title": "Polarization of the Vaccination Debate on Facebook"
}
| null | null | null | null | true | null |
17211
| null |
Default
| null | null |
null |
{
"abstract": " Personal electronic devices including smartphones give access to behavioural\nsignals that can be used to learn about the characteristics and preferences of\nindividuals. In this study, we explore the connection between demographic and\npsychological attributes and the digital behavioural records, for a cohort of\n7,633 people, closely representative of the US population with respect to\ngender, age, geographical distribution, education, and income. Along with the\ndemographic data, we collected self-reported assessments on validated\npsychometric questionnaires for moral traits and basic human values and\ncombined this information with passively collected multi-modal digital data\nfrom web browsing behaviour and smartphone usage. A machine learning framework\nwas then designed to infer both the demographic and psychological attributes\nfrom the behavioural data. In a cross-validated setting, our models predicted\ndemographic attributes with good accuracy as measured by the weighted AUROC\nscore (Area Under the Receiver Operating Characteristic), but were less\nperformant for the moral traits and human values. These results call for\nfurther investigation since they are still far from unveiling individuals'\npsychological fabric. This connection, along with the most predictive features\nthat we provide for each attribute, might prove useful for designing\npersonalised services, communication strategies, and interventions, and can be\nused to sketch a portrait of people with a similar worldview.\n",
"title": "Predicting Demographics, Moral Foundations, and Human Values from Digital Behaviors"
}
| null | null | null | null | true | null |
17212
| null |
Default
| null | null |
null |
{
"abstract": " We study the least squares regression function estimator over the class of\nreal-valued functions on $[0,1]^d$ that are increasing in each coordinate. For\nuniformly bounded signals and with a fixed, cubic lattice design, we establish\nthat the estimator achieves the minimax rate of order\n$n^{-\\min\\{2/(d+2),1/d\\}}$ in the empirical $L_2$ loss, up to poly-logarithmic\nfactors. Further, we prove a sharp oracle inequality, which reveals in\nparticular that when the true regression function is piecewise constant on $k$\nhyperrectangles, the least squares estimator enjoys a faster, adaptive rate of\nconvergence of $(k/n)^{\\min(1,2/d)}$, again up to poly-logarithmic factors.\nPrevious results are confined to the case $d \\leq 2$. Finally, we establish\ncorresponding bounds (which are new even in the case $d=2$) in the more\nchallenging random design setting. There are two surprising features of these\nresults: first, they demonstrate that it is possible for a global empirical\nrisk minimisation procedure to be rate optimal up to poly-logarithmic factors\neven when the corresponding entropy integral for the function class diverges\nrapidly; second, they indicate that the adaptation rate for shape-constrained\nestimators can be strictly worse than the parametric rate.\n",
"title": "Isotonic regression in general dimensions"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
17213
| null |
Validated
| null | null |
null |
{
"abstract": " The thermal stability of most electronic and photo-electronic devices\nstrongly depends on the relationship between Schottky Barrier Height (SBH) and\ntemperature. In this paper, the possible of thermionic current depicted via\ncorrect and reliability relationship between forward current and voltage is\nconsequently discussed, the intrinsic SBH insensitive to temperature can be\ncalculated by modification on Richardson- Dushman`s formula suggested in this\npaper. The results of application on four hetero-junctions prove that the\nmethod proposed is credible in this paper, this suggests that the I/V/T method\nis a feasible alternative to characterize these heterojunctions.\n",
"title": "Extraction of Schottky barrier height insensitive to temperature via forward currentvoltage- temperature measurements"
}
| null | null | null | null | true | null |
17214
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study emergent irreducible information in populations of\nrandomly generated computable systems that are networked and follow a\n\"Susceptible-Infected-Susceptible\" contagion model of imitation of the fittest\nneighbor. We show that there is a lower bound for the stationary prevalence (or\naverage density of \"infected\" nodes) that triggers an unlimited increase of the\nexpected local emergent algorithmic complexity (or information) of a node as\nthe population size grows. We call this phenomenon expected (local) emergent\nopen-endedness. In addition, we show that static networks with a power-law\ndegree distribution following the Barabási-Albert model satisfy this lower\nbound and, thus, display expected (local) emergent open-endedness.\n",
"title": "Emergent Open-Endedness from Contagion of the Fittest"
}
| null | null | null | null | true | null |
17215
| null |
Default
| null | null |
null |
{
"abstract": " This paper concerns the low Mach number limit of weak solutions to the\ncompressible Navier-Stokes equations for isentropic fluids in a bounded domain\nwith a Navier-slip boundary condition. In \\cite{DGLM99}, it has been proved\nthat if the velocity is imposed the homogeneous Dirichlet boundary condition,\nas the Mach number goes to 0, the velocity of the compressible flow converges\nstrongly in $L^2$ under the geometrical assumption (H) on the domain. We\njustify the same strong convergence when the slip length in the Navier\ncondition is the reciprocal of the square root of the Mach number.\n",
"title": "Incompressible Limit of isentropic Navier-Stokes equations with Navier-slip boundary"
}
| null | null | null | null | true | null |
17216
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, the formulas of some exponential sums over finite field,\nrelated to the Coulter's polynomial, are settled based on the Coulter's\ntheorems on Weil sums, which may have potential application in the construction\nof linear codes with few weights.\n",
"title": "On Some Exponential Sums Related to the Coulter's Polynomial"
}
| null | null | null | null | true | null |
17217
| null |
Default
| null | null |
null |
{
"abstract": " Preserving the privacy of individuals by protecting their sensitive\nattributes is an important consideration during microdata release. However, it\nis equally important to preserve the quality or utility of the data for at\nleast some targeted workloads. We propose a novel framework for privacy\npreservation based on the k-anonymity model that is ideally suited for\nworkloads that require preserving the probability distribution of the\nquasi-identifier variables in the data. Our framework combines the principles\nof distribution-preserving quantization and k-member clustering, and we\nspecialize it to two variants that respectively use intra-cluster and Gaussian\ndithering of cluster centers to achieve distribution preservation. We perform\ntheoretical analysis of the proposed schemes in terms of distribution\npreservation, and describe their utility in workloads such as covariate shift\nand transfer learning where such a property is necessary. Using extensive\nexperiments on real-world Medical Expenditure Panel Survey data, we demonstrate\nthe merits of our algorithms over standard k-anonymization for a hallmark\nhealth care application where an insurance company wishes to understand the\nrisk in entering a new market. Furthermore, by empirically quantifying the\nreidentification risk, we also show that the proposed approaches indeed\nmaintain k-anonymity.\n",
"title": "Distribution-Preserving k-Anonymity"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17218
| null |
Validated
| null | null |
null |
{
"abstract": " The interplay between superconductivity and charge density waves (CDW) in\n$H$-NbSe2 is not fully understood despite decades of study. Artificially\nintroduced disorder can tip the delicate balance between two competing forms of\nlong-range order, and reveal the underlying interactions that give rise to\nthem. Here we introduce disorders by electron irradiation and measure in-plane\nresistivity, Hall resistivity, X-ray scattering, and London penetration depth.\nWith increasing disorder, $T_{\\textrm{c}}$ varies nonmonotonically, whereas\n$T_{\\textrm{CDW}}$ monotonically decreases and becomes unresolvable above a\ncritical irradiation dose where $T_{\\textrm{c}}$ drops sharply. Our results\nimply that CDW order initially competes with superconductivity, but eventually\nassists it. We argue that at the transition where the long-range CDW order\ndisappears, the cooperation with superconductivity is dramatically suppressed.\nX-ray scattering and Hall resistivity measurements reveal that the short-range\nCDW survives above the transition. Superconductivity persists to much higher\ndose levels, consistent with fully gapped superconductivity and moderate\ninterband pairing.\n",
"title": "Using controlled disorder to probe the interplay between charge order and superconductivity in NbSe2"
}
| null | null | null | null | true | null |
17219
| null |
Default
| null | null |
null |
{
"abstract": " Background: The quality of a software product depends on the quality of the\nsoftware process followed in developing the product. Therefore, many higher\neducation institutions (HEI) and software organizations have implemented\nsoftware process improvement (SPI) training courses to improve the software\nquality. Objective: Because the duration of a course is a concern for HEI and\nsoftware organizations, we investigate whether the quality of software projects\nwill be improved by reorganizing the activities of the ten assignments of the\noriginal personal software process (PSP) course into a modified PSP having\nfewer assignments (i.e., seven assignments). Method: The assignments were\ndeveloped by following a modified PSP with fewer assignments but including the\nphases, forms, standards, and logs suggested in the original PSP. The\nmeasurement of the quality of the software assignments was based on defect\ndensity. Results: When the activities in the original PSP were reordered into\nfewer assignments, as practitioners progress through the PSP training, the\ndefect density improved with statistical significance. Conclusions: Our\nmodified PSP could be applied in academy and industrial environments which are\nconcerned in the sense of reducing the PSP training time\n",
"title": "A training process for improving the quality of software projects developed by a practitioner"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17220
| null |
Validated
| null | null |
null |
{
"abstract": " Although the Gaia catalogue on its own will be a very powerful tool, it is\nthe combination of this highly accurate archive with other archives that will\ntruly open up amazing possibilities for astronomical research. The advanced\ninteroperation of archives is based on cross-matching, leaving the user with\nthe feeling of working with one single data archive. The data retrieval should\nwork not only across data archives, but also across wavelength domains. The\nfirst step for seamless data access is the computation of the cross-match\nbetween Gaia and external surveys. The matching of astronomical catalogues is a\ncomplex and challenging problem both scientifically and technologically\n(especially when matching large surveys like Gaia). We describe the cross-match\nalgorithm used to pre-compute the match of Gaia Data Release 1 (DR1) with a\nselected list of large publicly available optical and IR surveys. The overall\nprinciples of the adopted cross-match algorithm are outlined. Details are given\non the developed algorithm, including the methods used to account for position\nerrors, proper motions, and environment; to define the neighbours; and to\ndefine the figure of merit used to select the most probable counterpart.\nStatistics on the results are also given. The results of the cross-match are\npart of the official Gaia DR1 catalogue.\n",
"title": "Gaia Data Release 1. Cross-match with external catalogues - Algorithm and results"
}
| null | null | null | null | true | null |
17221
| null |
Default
| null | null |
null |
{
"abstract": " We use 16 quarters of the \\textit{Kepler} mission data to analyze the transit\ntiming variations (TTVs) of the extrasolar planet Kepler-46b (KOI-872). Our\ndynamical fits confirm that the TTVs of this planet (period\n$P=33.648^{+0.004}_{-0.005}$ days) are produced by a non-transiting planet\nKepler-46c ($P=57.325^{+0.116}_{-0.098}$ days). The Bayesian inference tool\n\\texttt{MultiNest} is used to infer the dynamical parameters of Kepler-46b and\nKepler-46c. We find that the two planets have nearly coplanar and circular\norbits, with eccentricities $\\simeq 0.03$ somewhat higher than previously\nestimated. The masses of the two planets are found to be\n$M_{b}=0.885^{+0.374}_{-0.343}$ and $M_{c}=0.362^{+0.016}_{-0.016}$ Jupiter\nmasses, with $M_{b}$ being determined here from TTVs for the first time. Due to\nthe precession of its orbital plane, Kepler-46c should start transiting its\nhost star in a few decades from now.\n",
"title": "Masses of Kepler-46b, c from Transit Timing Variations"
}
| null | null | null | null | true | null |
17222
| null |
Default
| null | null |
null |
{
"abstract": " The reconstruction of water wave elevation from bottom pressure measurements\nis an important issue for coastal applications, but corresponds to a difficult\nmathematical problem. In this paper we present the derivation of a method which\nallows the elevation reconstruction of water waves in intermediate and shallow\nwaters. From comparisons with numerical Euler solutions and wave-tank\nexperiments we show that our nonlinear method provides much better results of\nthe surface elevation reconstruction compared to the linear transfer function\napproach commonly used in coastal applications. More specifically, our\nmethodaccurately reproduces the peaked and skewed shape of nonlinear wave\nfields. Therefore, it is particularly relevant for applications on extreme\nwaves and wave-induced sediment transport.\n",
"title": "Recovering water wave elevation from pressure measurements"
}
| null | null |
[
"Physics"
] | null | true | null |
17223
| null |
Validated
| null | null |
null |
{
"abstract": " The Schatten quasi-norm was introduced to bridge the gap between the trace\nnorm and rank function. However, existing algorithms are too slow or even\nimpractical for large-scale problems. Motivated by the equivalence relation\nbetween the trace norm and its bilinear spectral penalty, we define two\ntractable Schatten norms, i.e.\\ the bi-trace and tri-trace norms, and prove\nthat they are in essence the Schatten-$1/2$ and $1/3$ quasi-norms,\nrespectively. By applying the two defined Schatten quasi-norms to various rank\nminimization problems such as MC and RPCA, we only need to solve much smaller\nfactor matrices. We design two efficient linearized alternating minimization\nalgorithms to solve our problems and establish that each bounded sequence\ngenerated by our algorithms converges to a critical point. We also provide the\nrestricted strong convexity (RSC) based and MC error bounds for our algorithms.\nOur experimental results verified both the efficiency and effectiveness of our\nalgorithms compared with the state-of-the-art methods.\n",
"title": "Tractable and Scalable Schatten Quasi-Norm Approximations for Rank Minimization"
}
| null | null | null | null | true | null |
17224
| null |
Default
| null | null |
null |
{
"abstract": " Consider a truncated circular unitary matrix which is a $p_n$ by $p_n$\nsubmatrix of an $n$ by $n$ circular unitary matrix by deleting the last $n-p_n$\ncolumns and rows. Jiang and Qi (2017) proved that the maximum absolute value of\nthe eigenvalues (known as spectral radius) of the truncated matrix, after\nproperly normalized, converges in distribution to the Gumbel distribution if\n$p_n/n$ is bounded away from $0$ and $1$. In this paper we investigate the\nlimiting distribution of the spectral radius under one of the following four\nconditions: (1). $p_n\\to\\infty$ and $p_n/n\\to 0$ as $n\\to\\infty$; (2).\n$(n-p_n)/n\\to 0$ and $(n-p_n)/(\\log n)^3\\to\\infty$ as $n\\to\\infty$; (3).\n$n-p_n\\to\\infty$ and $(n-p_n)/\\log n\\to 0$ as $n\\to\\infty$ and (4). $n-p_n=k\\ge\n1$ is a fixed integer. We prove that the spectral radius converges in\ndistribution to the Gumbel distribution under the first three conditions and to\na reversed Weibull distribution under the fourth condition.\n",
"title": "Spectral Radii of Truncated Circular Unitary Matrices"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
17225
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, the task-related fMRI problem is treated in its matrix\nfactorization formulation. The focus of the reported work is on the dictionary\nlearning (DL) matrix factorization approach. A major novelty of the paper lies\nin the incorporation of well-established assumptions associated with the GLM\ntechnique, which is currently in use by the neuroscientists. These assumptions\nare embedded as constraints in the DL formulation. In this way, our approach\nprovides a framework of combining well-established and understood techniques\nwith a more ``modern'' and powerful tool. Furthermore, this paper offers a way\nto relax a major drawback associated with DL techniques; that is, the proper\ntuning of the DL regularization parameter. This parameter plays a critical role\nin DL-based fMRI analysis since it essentially determines the shape and\nstructures of the estimated functional brain networks. However, in actual fMRI\ndata analysis, the lack of ground truth renders the a priori choice of the\nregularization parameter a truly challenging task. Indeed, the values of the DL\nregularization parameter, associated with the $\\ell_1$ sparsity promoting norm,\ndo not convey any tangible physical meaning. So it is practically difficult to\nguess its proper value. In this paper, the DL problem is reformulated around a\nsparsity-promoting constraint that can directly be related to the minimum\namount of voxels that the spatial maps of the functional brain networks occupy.\nSuch information is documented and it is readily available to neuroscientists\nand experts in the field.\nThe proposed method is tested against a number of other popular techniques\nand the obtained performance gains are reported using a number of synthetic\nfMRI data. Results with real data have also been obtained in the context of a\nnumber of experiments and will be soon reported in a different publication.\n",
"title": "Information Assisted Dictionary Learning for fMRI data analysis"
}
| null | null |
[
"Statistics"
] | null | true | null |
17226
| null |
Validated
| null | null |
null |
{
"abstract": " Motivated by the task of clustering either $d$ variables or $d$ points into\n$K$ groups, we investigate efficient algorithms to solve the Peng-Wei (P-W)\n$K$-means semi-definite programming (SDP) relaxation. The P-W SDP has been\nshown in the literature to have good statistical properties in a variety of\nsettings, but remains intractable to solve in practice. To this end we propose\nFORCE, a new algorithm to solve this SDP relaxation. Compared to the naive\ninterior point method, our method reduces the computational complexity of\nsolving the SDP from $\\tilde{O}(d^7\\log\\epsilon^{-1})$ to\n$\\tilde{O}(d^{6}K^{-2}\\epsilon^{-1})$ arithmetic operations for an\n$\\epsilon$-optimal solution. Our method combines a primal first-order method\nwith a dual optimality certificate search, which when successful, allows for\nearly termination of the primal method. We show for certain variable clustering\nproblems that, with high probability, FORCE is guaranteed to find the optimal\nsolution to the SDP relaxation and provide a certificate of exact optimality.\nAs verified by our numerical experiments, this allows FORCE to solve the P-W\nSDP with dimensions in the hundreds in only tens of seconds. For a variation of\nthe P-W SDP where $K$ is not known a priori a slight modification of FORCE\nreduces the computational complexity of solving this problem as well: from\n$\\tilde{O}(d^7\\log\\epsilon^{-1})$ using a standard SDP solver to\n$\\tilde{O}(d^{4}\\epsilon^{-1})$.\n",
"title": "Efficient, Certifiably Optimal Clustering with Applications to Latent Variable Graphical Models"
}
| null | null |
[
"Statistics"
] | null | true | null |
17227
| null |
Validated
| null | null |
null |
{
"abstract": " We present wide-field (167 deg$^2$) weak lensing mass maps from the Hyper\nSupreme-Cam Subaru Strategic Program (HSC-SSP). We compare these weak lensing\nbased dark matter maps with maps of the distribution of the stellar mass\nassociated with luminous red galaxies. We find a strong correlation between\nthese two maps with a correlation coefficient of $\\rho=0.54\\pm0.03$ (for a\nsmoothing size of $8'$). This correlation is detected even with a smaller\nsmoothing scale of $2'$ ($\\rho=0.34\\pm 0.01$). This detection is made uniquely\npossible because of the high source density of the HSC-SSP weak lensing survey\n($\\bar{n}\\sim 25$ arcmin$^{-2}$). We also present a variety of tests to\ndemonstrate that our maps are not significantly affected by systematic effects.\nBy using the photometric redshift information associated with source galaxies,\nwe reconstruct a three-dimensional mass map. This three-dimensional mass map is\nalso found to correlate with the three-dimensional galaxy mass map.\nCross-correlation tests presented in this paper demonstrate that the HSC-SSP\nweak lensing mass maps are ready for further science analyses.\n",
"title": "Two- and three-dimensional wide-field weak lensing mass maps from the Hyper Suprime-Cam Subaru Strategic Program S16A data"
}
| null | null | null | null | true | null |
17228
| null |
Default
| null | null |
null |
{
"abstract": " Finite difference methods are traditionally used for modelling the time\ndomain in numerical weather prediction (NWP). Time-spectral solution is an\nattractive alternative for reasons of accuracy and efficiency and because time\nstep limitations associated with causal, CFL-like critera are avoided. In this\nwork, the Lorenz 1984 chaotic equations are solved using the time-spectral\nalgorithm GWRM. Comparisons of accuracy and efficiency are carried out for both\nexplicit and implicit time-stepping algorithms. It is found that the efficiency\nof the GWRM compares well with these methods, in particular at high accuracy.\nFor perturbative scenarios, the GWRM was found to be as much as four times\nfaster than the finite difference methods. A primary reason is that the GWRM\ntime intervals typically are two orders of magnitude larger than those of the\nfinite difference methods. The GWRM has the additional advantage to produce\nanalytical solutions in the form of Chebyshev series expansions. The results\nare encouraging for pursuing further studies, including spatial dependence, of\nthe relevance of time-spectral methods for NWP modelling.\n",
"title": "A Time-spectral Approach to Numerical Weather Prediction"
}
| null | null | null | null | true | null |
17229
| null |
Default
| null | null |
null |
{
"abstract": " Kinetic energy density functionals (KEDFs) are central to orbital-free\ndensity functional theory. Limitations on the spatial derivative dependencies\nof KEDFs have been claimed from differential virial theorems. We point out a\ncentral defect in the argument: the relationships are not true for an arbitrary\ndensity but hold only for the minimizing density and corresponding chemical\npotential. Contrary to the claims therefore, the relationships are not\nconstraints and provide no independent information about the spatial derivative\ndependencies of approximate KEDFs. A simple argument also shows that validity\nfor arbitrary $v$-representable densities is not restored by appeal to the\ndensity-potential bijection.\n",
"title": "Trivial Constraints on Orbital-free Kinetic Energy Density Functionals"
}
| null | null | null | null | true | null |
17230
| null |
Default
| null | null |
null |
{
"abstract": " The muti-layer information bottleneck (IB) problem, where information is\npropagated (or successively refined) from layer to layer, is considered. Based\non information forwarded by the preceding layer, each stage of the network is\nrequired to preserve a certain level of relevance with regards to a specific\nhidden variable, quantified by the mutual information. The hidden variables and\nthe source can be arbitrarily correlated. The optimal trade-off between rates\nof relevance and compression (or complexity) is obtained through a\nsingle-letter characterization, referred to as the rate-relevance region.\nConditions of successive refinabilty are given. Binary source with BSC hidden\nvariables and binary source with BSC/BEC mixed hidden variables are both proved\nto be successively refinable. We further extend our result to Guassian models.\nA counterexample of successive refinability is also provided.\n",
"title": "The Multi-layer Information Bottleneck Problem"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17231
| null |
Validated
| null | null |
null |
{
"abstract": " We propose a new mathematical model for $n-k$-dimensional non-linear\ncorrelations with intrinsic scatter in $n$-dimensional data. The model is based\non Riemannian geometry, and is naturally symmetric with respect to the measured\nvariables and invariant under coordinate transformations. We combine the model\nwith a Bayesian approach for estimating the parameters of the correlation\nrelation and the intrinsic scatter. A side benefit of the approach is that\ncensored and truncated datasets and independent, arbitrary measurement errors\ncan be incorporated. We also derive analytic likelihoods for the typical\nastrophysical use case of linear relations in $n$-dimensional Euclidean space.\nWe pay particular attention to the case of linear regression in two dimensions,\nand compare our results to existing methods. Finally, we apply our methodology\nto the well-known $M_\\text{BH}$-$\\sigma$ correlation between the mass of a\nsupermassive black hole in the centre of a galactic bulge and the corresponding\nbulge velocity dispersion. The main result of our analysis is that the most\nlikely slope of this correlation is $\\sim 6$ for the datasets used, rather than\nthe values in the range $\\sim 4$-$5$ typically quoted in the literature for\nthese data.\n",
"title": "A geometric approach to non-linear correlations with intrinsic scatter"
}
| null | null |
[
"Physics",
"Mathematics",
"Statistics"
] | null | true | null |
17232
| null |
Validated
| null | null |
null |
{
"abstract": " A central problem of algebraic topology is to understand the homotopy groups\n$\\pi_d(X)$ of a topological space $X$. For the computational version of the\nproblem, it is well known that there is no algorithm to decide whether the\nfundamental group $\\pi_1(X)$ of a given finite simplicial complex $X$ is\ntrivial. On the other hand, there are several algorithms that, given a finite\nsimplicial complex $X$ that is simply connected (i.e., with $\\pi_1(X)$\ntrivial), compute the higher homotopy group $\\pi_d(X)$ for any given $d\\geq 2$.\n%The first such algorithm was given by Brown, and more recently, Čadek et\nal.\nHowever, these algorithms come with a caveat: They compute the isomorphism\ntype of $\\pi_d(X)$, $d\\geq 2$ as an \\emph{abstract} finitely generated abelian\ngroup given by generators and relations, but they work with very implicit\nrepresentations of the elements of $\\pi_d(X)$. Converting elements of this\nabstract group into explicit geometric maps from the $d$-dimensional sphere\n$S^d$ to $X$ has been one of the main unsolved problems in the emerging field\nof computational homotopy theory.\nHere we present an algorithm that, given a~simply connected space $X$,\ncomputes $\\pi_d(X)$ and represents its elements as simplicial maps from a\nsuitable triangulation of the $d$-sphere $S^d$ to $X$. For fixed $d$, the\nalgorithm runs in time exponential in $size(X)$, the number of simplices of\n$X$. Moreover, we prove that this is optimal: For every fixed $d\\geq 2$, we\nconstruct a family of simply connected spaces $X$ such that for any simplicial\nmap representing a generator of $\\pi_d(X)$, the size of the triangulation of\n$S^d$ on which the map is defined, is exponential in $size(X)$.\n",
"title": "Computing simplicial representatives of homotopy group elements"
}
| null | null | null | null | true | null |
17233
| null |
Default
| null | null |
null |
{
"abstract": " Optimization is becoming a crucial element in industrial applications\ninvolving sustainable alternative energy systems. During the design of such\nsystems, the engineer/decision maker would often encounter noise factors (e.g.\nsolar insolation and ambient temperature fluctuations) when their system\ninteracts with the environment. In this chapter, the sizing and design\noptimization of the solar powered irrigation system was considered. This\nproblem is multivariate, noisy, nonlinear and multiobjective. This design\nproblem was tackled by first using the Fuzzy Type II approach to model the\nnoise factors. Consequently, the Bacterial Foraging Algorithm (BFA) (in the\ncontext of a weighted sum framework) was employed to solve this multiobjective\nfuzzy design problem. This method was then used to construct the approximate\nPareto frontier as well as to identify the best solution option in a fuzzy\nsetting. Comprehensive analyses and discussions were performed on the generated\nnumerical results with respect to the implemented solution methods.\n",
"title": "Multiobjective Optimization of Solar Powered Irrigation System with Fuzzy Type-2 Noise Modelling"
}
| null | null | null | null | true | null |
17234
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study convergence properties of the gradient\nExpectation-Maximization algorithm \\cite{lange1995gradient} for Gaussian\nMixture Models for general number of clusters and mixing coefficients. We\nderive the convergence rate depending on the mixing coefficients, minimum and\nmaximum pairwise distances between the true centers and dimensionality and\nnumber of components; and obtain a near-optimal local contraction radius. While\nthere have been some recent notable works that derive local convergence rates\nfor EM in the two equal mixture symmetric GMM, in the more general case, the\nderivations need structurally different and non-trivial arguments. We use\nrecent tools from learning theory and empirical processes to achieve our\ntheoretical results.\n",
"title": "Convergence Analysis of Gradient EM for Multi-component Gaussian Mixture"
}
| null | null | null | null | true | null |
17235
| null |
Default
| null | null |
null |
{
"abstract": " The direct detection of gravitational wave by Laser Interferometer\nGravitational-Wave Observatory indicates the coming of the era of\ngravitational-wave astronomy and gravitational-wave cosmology. It is expected\nthat more and more gravitational-wave events will be detected by currently\nexisting and planned gravitational-wave detectors. The gravitational waves open\na new window to explore the Universe and various mysteries will be disclosed\nthrough the gravitational-wave detection, combined with other cosmological\nprobes. The gravitational-wave physics is not only related to gravitation\ntheory, but also is closely tied to fundamental physics, cosmology and\nastrophysics. In this review article, three kinds of sources of gravitational\nwaves and relevant physics will be discussed, namely gravitational waves\nproduced during the inflation and preheating phases of the Universe, the\ngravitational waves produced during the first-order phase transition as the\nUniverse cools down and the gravitational waves from the three phases:\ninspiral, merger and ringdown of a compact binary system, respectively. We will\nalso discuss the gravitational waves as a standard siren to explore the\nevolution of the Universe.\n",
"title": "The Gravitational-Wave Physics"
}
| null | null |
[
"Physics"
] | null | true | null |
17236
| null |
Validated
| null | null |
null |
{
"abstract": " Approximations during program analysis are a necessary evil, as they ensure\nessential properties, such as soundness and termination of the analysis, but\nthey also imply not always producing useful results. Automatic techniques have\nbeen studied to prevent precision loss, typically at the expense of larger\nresource consumption. In both cases (i.e., when analysis produces inaccurate\nresults and when resource consumption is too high), it is necessary to have\nsome means for users to provide information to guide analysis and thus improve\nprecision and/or performance. We present techniques for supporting within an\nabstract interpretation framework a rich set of assertions that can deal with\nmultivariance/context-sensitivity, and can handle different run-time semantics\nfor those assertions that cannot be discharged at compile time. We show how the\nproposed approach can be applied to both improving precision and accelerating\nanalysis. We also provide some formal results on the effects of such assertions\non the analysis results.\n",
"title": "Multivariant Assertion-based Guidance in Abstract Interpretation"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17237
| null |
Validated
| null | null |
null |
{
"abstract": " Through the development of efficient algorithms, data structures and\npreprocessing techniques, real-world shortest path problems in street networks\nare now very fast to solve. But in reality, the exact travel times along each\narc in the network may not be known. This lead to the development of robust\nshortest path problems, where all possible arc travel times are contained in a\nso-called uncertainty set of possible outcomes.\nResearch in robust shortest path problems typically assumes this set to be\ngiven, and provides complexity results as well as algorithms depending on its\nshape. However, what can actually be observed in real-world problems are only\ndiscrete raw data points. The shape of the uncertainty is already a modelling\nassumption. In this paper we test several of the most widely used assumptions\non the uncertainty set using real-world traffic measurements provided by the\nCity of Chicago. We calculate the resulting different robust solutions, and\nevaluate which uncertainty approach is actually reasonable for our data. This\nanchors theoretical research in a real-world application and allows us to point\nout which robust models should be the future focus of algorithmic development.\n",
"title": "An Experimental Comparison of Uncertainty Sets for Robust Shortest Path Problems"
}
| null | null | null | null | true | null |
17238
| null |
Default
| null | null |
null |
{
"abstract": " Nonlinear wave interactions affect the evolution of steep wave groups, their\nbreaking and the associated kinematic field. Laboratory experiments are\nperformed to investigate the effect of the underlying focussing mechanism on\nthe shape of the breaking wave and its velocity field. In this regard, it is\nfound that the shape of the wave spectrum plays a substantial role. Broader\nunderlying wave spectra leads to energetic plungers at a relatively low\namplitude. For narrower spectra waves break at a higher amplitudes but with a\nless energetic spiller. Comparison with standard engineering methods commonly\nused to predict the velocity underneath extreme waves shows that, under certain\nconditions, the measured velocity profile strongly deviates from engineering\npredictions.\n",
"title": "An experimental comparison of velocities underneath focussed breaking waves"
}
| null | null | null | null | true | null |
17239
| null |
Default
| null | null |
null |
{
"abstract": " The single-particle spectral function measures the density of electronic\nstates (DOS) in a material as a function of both momentum and energy, providing\ncentral insights into phenomena such as superconductivity and Mott insulators.\nWhile scanning tunneling microscopy (STM) and other tunneling methods have\nprovided partial spectral information, until now only angle-resolved\nphotoemission spectroscopy (ARPES) has permitted a comprehensive determination\nof the spectral function of materials in both momentum and energy. However,\nARPES operates only on electronic systems at the material surface and cannot\nwork in the presence of applied magnetic fields. Here, we demonstrate a new\nmethod for determining the full momentum and energy resolved electronic\nspectral function of a two-dimensional (2D) electronic system embedded in a\nsemiconductor. In contrast with ARPES, the technique remains operational in the\npresence of large externally applied magnetic fields and functions for\nelectronic systems with zero electrical conductivity or with zero electron\ndensity. It provides a direct high-resolution and high-fidelity probe of the\ndispersion and dynamics of the interacting 2D electron system. By ensuring the\nsystem of interest remains under equilibrium conditions, we uncover delicate\nsignatures of many-body effects involving electron-phonon interactions,\nplasmons, polarons, and a novel phonon analog of the vacuum Rabi splitting in\natomic systems.\n",
"title": "Full Momentum and Energy Resolved Spectral Function of a 2D Electronic System"
}
| null | null | null | null | true | null |
17240
| null |
Default
| null | null |
null |
{
"abstract": " The purpose of this article is to determine explicitly the complete surfaces\nwith parallel mean curvature vector, both in the complex projective plane and\nthe complex hyperbolic plane. The main results are as follows: When the\ncurvature of the ambient space is positive, there exists a unique such surface\nup to rigid motions of the target space. On the other hand, when the curvature\nof the ambient space is negative, there are `non-trivial' complete parallel\nmean curvature surfaces generated by Jacobi elliptic functions and they exhaust\nsuch surfaces.\n",
"title": "Complete parallel mean curvature surfaces in two-dimensional complex space-forms"
}
| null | null | null | null | true | null |
17241
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we introduce a new type of linear classifier that is\nimplemented in a chemical form. We propose a novel encoding technique which\nsimultaneously represents multiple datasets in an array of microliter-scale\nchemical mixtures. Parallel computations on these datasets are performed as\nrobotic liquid handling sequences, whose outputs are analyzed by\nhigh-performance liquid chromatography. As a proof of concept, we chemically\nencode several MNIST images of handwritten digits and demonstrate successful\nchemical-domain classification of the digits using volumetric perceptrons. We\nadditionally quantify the performance of our method with a larger dataset of\nbinary vectors and compare the experimental measurements against predicted\nresults. Paired with appropriate chemical analysis tools, our approach can work\non increasingly parallel datasets. We anticipate that related approaches will\nbe scalable to multilayer neural networks and other more complex algorithms.\nMuch like recent demonstrations of archival data storage in DNA, this work\nblurs the line between chemical and electrical information systems, and offers\nearly insight into the computational efficiency and massive parallelism which\nmay come with computing in chemical domains.\n",
"title": "Parallelized Linear Classification with Volumetric Chemical Perceptrons"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
17242
| null |
Validated
| null | null |
null |
{
"abstract": " Multi-label image classification is a fundamental but challenging task in\ncomputer vision. Great progress has been achieved by exploiting semantic\nrelations between labels in recent years. However, conventional approaches are\nunable to model the underlying spatial relations between labels in multi-label\nimages, because spatial annotations of the labels are generally not provided.\nIn this paper, we propose a unified deep neural network that exploits both\nsemantic and spatial relations between labels with only image-level\nsupervisions. Given a multi-label image, our proposed Spatial Regularization\nNetwork (SRN) generates attention maps for all labels and captures the\nunderlying relations between them via learnable convolutions. By aggregating\nthe regularized classification results with original results by a ResNet-101\nnetwork, the classification performance can be consistently improved. The whole\ndeep neural network is trained end-to-end with only image-level annotations,\nthus requires no additional efforts on image annotations. Extensive evaluations\non 3 public datasets with different types of labels show that our approach\nsignificantly outperforms state-of-the-arts and has strong generalization\ncapability. Analysis of the learned SRN model demonstrates that it can\neffectively capture both semantic and spatial relations of labels for improving\nclassification performance.\n",
"title": "Learning Spatial Regularization with Image-level Supervisions for Multi-label Image Classification"
}
| null | null | null | null | true | null |
17243
| null |
Default
| null | null |
null |
{
"abstract": " The atmospheres of between one quarter and one half of observed single white\ndwarfs in the Milky Way contain heavy element pollution from planetary debris.\nThe pollution observed in white dwarfs in binary star systems is, however, less\nclear, because companion star winds can generate a stream of matter which is\naccreted by the white dwarf. Here we (i) discuss the necessity or lack thereof\nof a major planet in order to pollute a white dwarf with orbiting minor planets\nin both single and binary systems, and (ii) determine the critical binary\nseparation beyond which the accretion source is from a planetary system. We\nhence obtain user-friendly functions relating this distance to the masses and\nradii of both stars, the companion wind, and the accretion rate onto the white\ndwarf, for a wide variety of published accretion prescriptions. We find that\nfor the majority of white dwarfs in known binaries, if pollution is detected,\nthen that pollution should originate from planetary material.\n",
"title": "The critical binary star separation for a planetary system origin of white dwarf pollution"
}
| null | null | null | null | true | null |
17244
| null |
Default
| null | null |
null |
{
"abstract": " We present a quantization of an isomorphism of Mirković and Vybornov which\nrelates the intersection of a Slodowy slice and a nilpotent orbit closure in\n$\\mathfrak{gl}_N$ , to a slice between spherical Schubert varieties in the\naffine Grassmannian of $PGL_n$ (with weights encoded by the Jordan types of the\nnilpotent orbits). A quantization of the former variety is provided by a\nparabolic W-algebra and of the latter by a truncated shifted Yangian. Building\non earlier work of Brundan and Kleshchev, we define an explicit isomorphism\nbetween these non-commutative algebras, and show that its classical limit is a\nvariation of the original isomorphism of Mirković and Vybornov. As a\ncorollary, we deduce that the W-algebra is free as a left (or right) module\nover its Gelfand-Tsetlin subalgebra, as conjectured by Futorny, Molev, and\nOvsienko.\n",
"title": "A quantum Mirković-Vybornov isomorphism"
}
| null | null |
[
"Mathematics"
] | null | true | null |
17245
| null |
Validated
| null | null |
null |
{
"abstract": " This paper is concerned with a multi-asset mean-variance portfolio selection\nproblem under model uncertainty. We develop a continuous time framework for\ntaking into account ambiguity aversion about both expected return rates and\ncorrelation matrix of the assets, and for studying the effects on portfolio\ndiversification. We prove a separation principle for the associated robust\ncontrol problem, which allows to reduce the determination of the optimal\ndynamic strategy to the parametric computation of the minimal risk premium\nfunction. Our results provide a justification for under-diversification, as\ndocumented in empirical studies. We explicitly quantify the degree of\nunder-diversification in terms of correlation and Sharpe ratio ambiguity. In\nparticular, we show that an investor with a poor confidence in the expected\nreturn estimation does not hold any risky asset, and on the other hand, trades\nonly one risky asset when the level of ambiguity on correlation matrix is\nlarge. This extends to the continuous-time setting the results obtained by\nGarlappi, Uppal and Wang [13], and Liu and Zeng [24] in a one-period model. JEL\nClassification: G11, C61 MSC Classification: 91G10, 91G80, 60H30\n",
"title": "Portfolio diversification and model uncertainty: a robust dynamic mean-variance approach"
}
| null | null |
[
"Quantitative Finance"
] | null | true | null |
17246
| null |
Validated
| null | null |
null |
{
"abstract": " Deep learning has enabled major advances in the fields of computer vision,\nnatural language processing, and multimedia among many others. Developing a\ndeep learning system is arduous and complex, as it involves constructing neural\nnetwork architectures, managing training/trained models, tuning optimization\nprocess, preprocessing and organizing data, etc. TensorLayer is a versatile\nPython library that aims at helping researchers and engineers efficiently\ndevelop deep learning systems. It offers rich abstractions for neural networks,\nmodel and data management, and parallel workflow mechanism. While boosting\nefficiency, TensorLayer maintains both performance and scalability. TensorLayer\nwas released in September 2016 on GitHub, and has helped people from academia\nand industry develop real-world applications of deep learning.\n",
"title": "TensorLayer: A Versatile Library for Efficient Deep Learning Development"
}
| null | null | null | null | true | null |
17247
| null |
Default
| null | null |
null |
{
"abstract": " Undesired unintentional doping and doping limits in semiconductors are\ntypically caused by compensating defects with low formation energies. Since the\nformation energy of a charged defect depends linearly on the Fermi level,\ndoping limits can be especially pronounced in wide bandgap semiconductors where\nthe Fermi level can vary substantially. Introduction of non-equilibrium carrier\nconcentrations during growth or processing alters the chemical potentials of\nband carriers and thus provides the possibility of modifying populations of\ncharged defects in ways impossible at thermal equilibrium. Herein we\ndemonstrate that, for an ergodic system with excess carriers, the rates of\ncarrier capture and emission involving a defect charge transition level\nrigorously determine the admixture of electron and hole quasi-Fermi levels\ndetermining the formation energy of non-zero charge states of that defect type.\nTo catalog the range of possible responses to excess carriers, we investigate\nthe behavior of a single donor-like defect as functions of extrinsic doping and\nenergy of the charge transition level. The technologically most important\nfinding is that excess carriers will increase the formation energy of\ncompensating defects for most values of the charge transition level in the\nbandgap. Thus, it may be possible to overcome limitations on doping imposed by\nnative defects. Cases also exist in wide bandgap semiconductors in which the\nconcentration of defects with the same charge polarity as the majority dopant\nis either left unchanged or actually increases. The causes of these various\nbehaviors are rationalized in terms of the capture and emission rates and\nguidelines for carrying out experimental tests of this model are given.\n",
"title": "Effects of excess carriers on native defects in wide bandgap semiconductors: illumination as a method to enhance p-type doping"
}
| null | null | null | null | true | null |
17248
| null |
Default
| null | null |
null |
{
"abstract": " In recent years, many research works propose to embed the network structured\ndata into a low-dimensional feature space, where each node is represented as a\nfeature vector. However, due to the detachment of embedding process with\nexternal tasks, the learned embedding results by most existing embedding models\ncan be ineffective for application tasks with specific objectives, e.g.,\ncommunity detection or information diffusion. In this paper, we propose study\nthe application oriented heterogeneous social network embedding problem.\nSignificantly different from the existing works, besides the network structure\npreservation, the problem should also incorporate the objectives of external\napplications in the objective function. To resolve the problem, in this paper,\nwe propose a novel network embedding framework, namely the \"appLicAtion\norienTed neTwork Embedding\" (Latte) model. In Latte, the heterogeneous network\nstructure can be applied to compute the node \"diffusive proximity\" scores,\nwhich capture both local and global network structures. Based on these computed\nscores, Latte learns the network representation feature vectors by extending\nthe autoencoder model model to the heterogeneous network scenario, which can\nalso effectively unite the objectives of network embedding and external\napplication tasks. Extensive experiments have been done on real-world\nheterogeneous social network datasets, and the experimental results have\ndemonstrated the outstanding performance of Latte in learning the\nrepresentation vectors for specific application tasks.\n",
"title": "LATTE: Application Oriented Social Network Embedding"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17249
| null |
Validated
| null | null |
null |
{
"abstract": " For applications exploiting the valley pseudospin degree of freedom in\ntransition metal dichalcogenide monolayers, efficient preparation of electrons\nor holes in a single valley is essential. Here, we show that a magnetic field\nof 7 Tesla leads to a near-complete valley polarization of electrons in MoSe2\nmonolayer with a density 1.6x10^{12} cm^{-2}; in the absence of exchange\ninteractions favoring single-valley occupancy, a similar degree of valley\npolarization would have required a pseudospin g-factor exceeding 40. To\ninvestigate the magnetic response, we use polarization resolved\nphotoluminescence as well as resonant reflection measurements. In the latter,\nwe observe gate voltage dependent transfer of oscillator strength from the\nexciton to the attractive-Fermi-polaron: stark differences in the spectrum of\nthe two light helicities provide a confirmation of valley polarization. Our\nfindings suggest an interaction induced giant paramagnetic response of MoSe2,\nwhich paves the way for valleytronics applications.\n",
"title": "Giant paramagnetism induced valley polarization of electrons in charge-tunable monolayer MoSe2"
}
| null | null | null | null | true | null |
17250
| null |
Default
| null | null |
null |
{
"abstract": " Predicting highrisk vascular diseases is a significant issue in the medical\ndomain. Most predicting methods predict the prognosis of patients from\npathological and radiological measurements, which are expensive and require\nmuch time to be analyzed. Here we propose deep attention models that predict\nthe onset of the high risky vascular disease from symbolic medical histories\nsequence of hypertension patients such as ICD-10 and pharmacy codes only,\nMedical History-based Prediction using Attention Network (MeHPAN). We\ndemonstrate two types of attention models based on 1) bidirectional gated\nrecurrent unit (R-MeHPAN) and 2) 1D convolutional multilayer model (C-MeHPAN).\nTwo MeHPAN models are evaluated on approximately 50,000 hypertension patients\nwith respect to precision, recall, f1-measure and area under the curve (AUC).\nExperimental results show that our MeHPAN methods outperform standard\nclassification models. Comparing two MeHPANs, R-MeHPAN provides more better\ndiscriminative capability with respect to all metrics while C-MeHPAN presents\nmuch shorter training time with competitive accuracy.\n",
"title": "Highrisk Prediction from Electronic Medical Records via Deep Attention Networks"
}
| null | null | null | null | true | null |
17251
| null |
Default
| null | null |
null |
{
"abstract": " Understanding the evolution of human society, as a complex adaptive system,\nis a task that has been looked upon from various angles. In this paper, we\nsimulate an agent-based model with a high enough population tractably. To do\nthis, we characterize an entity called \\textit{society}, which helps us reduce\nthe complexity of each step from $\\mathcal{O}(n^2)$ to $\\mathcal{O}(n)$. We\npropose a very realistic setting, where we design a joint alternate\nmaximization step algorithm to maximize a certain \\textit{fitness} function,\nwhich we believe simulates the way societies develop. Our key contributions\ninclude (i) proposing a novel protocol for simulating the evolution of a\nsociety with cheap, non-optimal joint alternate maximization steps (ii)\nproviding a framework for carrying out experiments that adhere to this\njoint-optimization simulation framework (iii) carrying out experiments to show\nthat it makes sense empirically (iv) providing an alternate justification for\nthe use of \\textit{society} in the simulations.\n",
"title": "Agent based simulation of the evolution of society as an alternate maximization problem"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17252
| null |
Validated
| null | null |
null |
{
"abstract": " Autonomic nervous system (ANS) activity is altered in autism spectrum\ndisorder (ASD). Heart rate variability (HRV) derived from electrocardiogram\n(ECG) has been a powerful tool to identify alterations in ANS due to a plethora\nof pathophysiological conditions, including psychological ones such as\ndepression. ECG-derived HRV thus carries a yet to be explored potential to be\nused as a diagnostic and follow-up biomarker of ASD. However, few studies have\nexplored this potential. In a cohort of boys (ages 8 - 11 years) with (n=18)\nand without ASD (n=18), we tested a set of linear and nonlinear HRV measures,\nincluding phase rectified signal averaging (PRSA), applied to a segment of ECG\ncollected under resting conditions for their predictive properties of ASD. We\nidentified HRV measures derived from time, frequency and geometric\nsignal-analytical domains which are changed in ASD children relative to peers\nwithout ASD and correlate to psychometric scores (p<0.05 for each). Receiver\noperating curves area ranged between 0.71 - 0.74 for each HRV measure. Despite\nbeing a small cohort lacking external validation, these promising preliminary\nresults warrant larger prospective validation studies.\n",
"title": "Can a heart rate variability biomarker identify the presence of autism spectrum disorder in eight year old children?"
}
| null | null | null | null | true | null |
17253
| null |
Default
| null | null |
null |
{
"abstract": " Unsupervised learning of low-dimensional, semantic representations of words\nand entities has recently gained attention. In this paper we describe the\nSemantic Entity Retrieval Toolkit (SERT) that provides implementations of our\npreviously published entity representation models. The toolkit provides a\nunified interface to different representation learning algorithms, fine-grained\nparsing configuration and can be used transparently with GPUs. In addition,\nusers can easily modify existing models or implement their own models in the\nframework. After model training, SERT can be used to rank entities according to\na textual query and extract the learned entity/word representation for use in\ndownstream algorithms, such as clustering or recommendation.\n",
"title": "Semantic Entity Retrieval Toolkit"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17254
| null |
Validated
| null | null |
null |
{
"abstract": " In the Best-$k$-Arm problem, we are given $n$ stochastic bandit arms, each\nassociated with an unknown reward distribution. We are required to identify the\n$k$ arms with the largest means by taking as few samples as possible. In this\npaper, we make progress towards a complete characterization of the\ninstance-wise sample complexity bounds for the Best-$k$-Arm problem. On the\nlower bound side, we obtain a novel complexity term to measure the sample\ncomplexity that every Best-$k$-Arm instance requires. This is derived by an\ninteresting and nontrivial reduction from the Best-$1$-Arm problem. We also\nprovide an elimination-based algorithm that matches the instance-wise lower\nbound within doubly-logarithmic factors. The sample complexity of our algorithm\nstrictly dominates the state-of-the-art for Best-$k$-Arm (module constant\nfactors).\n",
"title": "Nearly Instance Optimal Sample Complexity Bounds for Top-k Arm Selection"
}
| null | null | null | null | true | null |
17255
| null |
Default
| null | null |
null |
{
"abstract": " We study equilibrium measures (Käenmäki measures) supported on\nself-affine sets generated by a finite collection of diagonal and anti-diagonal\nmatrices acting on the plane and satisfying the strong separation property. Our\nmain result is that such measures are exact dimensional and the dimension\nsatisfies the Ledrappier-Young formula, which gives an explicit expression for\nthe dimension in terms of the entropy and Lyapunov exponents as well as the\ndimension of the important coordinate projection of the measure. In particular,\nwe do this by showing that the Käenmäki measure is equal to the sum of (the\npushforwards) of two Gibbs measures on an associated subshift of finite type.\n",
"title": "Dimensions of equilibrium measures on a class of planar self-affine sets"
}
| null | null | null | null | true | null |
17256
| null |
Default
| null | null |
null |
{
"abstract": " We present a thermal emission spectrum of the bloated hot Jupiter HAT-P-32Ab\nfrom a single eclipse observation made in spatial scan mode with the Wide Field\nCamera 3 (WFC3) aboard the Hubble Space Telescope (HST). The spectrum covers\nthe wavelength regime from 1.123 to 1.644 microns which is binned into 14\neclipse depths measured to an averaged precision of 104 parts-per million. The\nspectrum is unaffected by a dilution from the close M-dwarf companion\nHAT-P-32B, which was fully resolved. We complemented our spectrum with\nliterature results and performed a comparative forward and retrieval analysis\nwith the 1D radiative-convective ATMO model. Assuming solar abundance of the\nplanet atmosphere, we find that the measured spectrum can best be explained by\nthe spectrum of a blackbody isothermal atmosphere with Tp = 1995 +/- 17K, but\ncan equally-well be described by a spectrum with modest thermal inversion. The\nretrieved spectrum suggests emission from VO at the WFC3 wavelengths and no\nevidence of the 1.4 micron water feature. The emission models with temperature\nprofiles decreasing with height are rejected at a high confidence. An\nisothermal or inverted spectrum can imply a clear atmosphere with an absorber,\na dusty cloud deck or a combination of both. We find that the planet can have\ncontinuum of values for the albedo and recirculation, ranging from high albedo\nand poor recirculation to low albedo and efficient recirculation. Optical\nspectroscopy of the planet's day-side or thermal emission phase curves can\npotentially resolve the current albedo with recirculation degeneracy.\n",
"title": "Hubble PanCET: An isothermal day-side atmosphere for the bloated gas-giant HAT-P-32Ab"
}
| null | null |
[
"Physics"
] | null | true | null |
17257
| null |
Validated
| null | null |
null |
{
"abstract": " Deep learning yields great results across many fields, from speech\nrecognition, image classification, to translation. But for each problem,\ngetting a deep model to work well involves research into the architecture and a\nlong period of tuning. We present a single model that yields good results on a\nnumber of problems spanning multiple domains. In particular, this single model\nis trained concurrently on ImageNet, multiple translation tasks, image\ncaptioning (COCO dataset), a speech recognition corpus, and an English parsing\ntask. Our model architecture incorporates building blocks from multiple\ndomains. It contains convolutional layers, an attention mechanism, and\nsparsely-gated layers. Each of these computational blocks is crucial for a\nsubset of the tasks we train on. Interestingly, even if a block is not crucial\nfor a task, we observe that adding it never hurts performance and in most cases\nimproves it on all tasks. We also show that tasks with less data benefit\nlargely from joint training with other tasks, while performance on large tasks\ndegrades only slightly if at all.\n",
"title": "One Model To Learn Them All"
}
| null | null | null | null | true | null |
17258
| null |
Default
| null | null |
null |
{
"abstract": " Neural networks have been used prominently in several machine learning and\nstatistics applications. In general, the underlying optimization of neural\nnetworks is non-convex which makes their performance analysis challenging. In\nthis paper, we take a novel approach to this problem by asking whether one can\nconstrain neural network weights to make its optimization landscape have good\ntheoretical properties while at the same time, be a good approximation for the\nunconstrained one. For two-layer neural networks, we provide affirmative\nanswers to these questions by introducing Porcupine Neural Networks (PNNs)\nwhose weight vectors are constrained to lie over a finite set of lines. We show\nthat most local optima of PNN optimizations are global while we have a\ncharacterization of regions where bad local optimizers may exist. Moreover, our\ntheoretical and empirical results suggest that an unconstrained neural network\ncan be approximated using a polynomially-large PNN.\n",
"title": "Porcupine Neural Networks: (Almost) All Local Optima are Global"
}
| null | null | null | null | true | null |
17259
| null |
Default
| null | null |
null |
{
"abstract": " Precise knowledge of the static density response function (SDRF) of the\nuniform electron gas (UEG) serves as key input for numerous applications, most\nimportantly for density functional theory beyond generalized gradient\napproximations. Here we extend the configuration path integral Monte Carlo\n(CPIMC) formalism that was previously applied to the spatially uniform electron\ngas to the case of an inhomogeneous electron gas by adding a spatially periodic\nexternal potential. This procedure has recently been successfully used in\npermutation blocking path integral Monte Carlo simulations (PB-PIMC) of the\nwarm dense electron gas [Dornheim \\textit{et al.}, Phys. Rev. E in press,\narXiv:1706.00315], but this method is restricted to low and moderate densities.\nImplementing this procedure into CPIMC allows us to obtain exact finite\ntemperature results for the SDRF of the electron gas at \\textit{high to\nmoderate densities} closing the gap left open by the PB-PIMC data. In this\npaper we demonstrate how the CPIMC formalism can be efficiently extended to the\nspatially inhomogeneous electron gas and present the first data points.\nFinally, we discuss finite size errors involved in the quantum Monte Carlo\nresults for the SDRF in detail and present a solution how to remove them that\nis based on a generalization of ground state techniques.\n",
"title": "Configuration Path Integral Monte Carlo Approach to the Static Density Response of the Warm Dense Electron Gas"
}
| null | null | null | null | true | null |
17260
| null |
Default
| null | null |
null |
{
"abstract": " The magnetocrystalline anisotropy exhibited in PrPd$_2$Ge$_2$ single crystal\nhas been investigated by measuring the magnetization, magnetic susceptibility,\nelectrical resistivity and heat capacity. PrPd$_2$Ge$_2$ crystallizes in the\nwell known ThCr$_2$Si$_2$\\--type tetragonal structure. The antiferromagnetic\nordering is confirmed as 5.1~K with the [001]-axis as the easy axis of\nmagnetization. A superzone gap formation is observed from the electrical\nresistivity measurement when the current is passed along the [001] direction.\nThe crystal electric field (CEF) analysis on the magnetic susceptibility,\nmagnetization and the heat capacity measurements confirms a doublet ground\nstate with a relatively low over all CEF level splitting. The CEF level\nspacings and the Zeeman splitting at high fields become comparable and lead to\nmetamagnetic transition at 34~T due to the CEF level crossing.\n",
"title": "Superzone gap formation and low lying crystal electric field levels in PrPd$_2$Ge$_2$ single crystal"
}
| null | null |
[
"Physics"
] | null | true | null |
17261
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, we experimentally demonstrate a real-time software defined\nmultiple input multiple output (MIMO) visible light communication (VLC) system\nemploying link adaptation of spatial multiplexing and spatial diversity.\nReal-time MIMO signal processing is implemented by using the Field Programmable\nGate Array (FPGA) based Universal Software Radio Peripheral (USRP) devices.\nSoftware defined implantation of MIMO VLC can assist in enabling an adaptive\nand reconfigurable communication system without hardware changes. We measured\nthe error vector magnitude (EVM), bit error rate (BER) and spectral efficiency\nperformance for single carrier M-QAM MIMO VLC using spatial diversity and\nspatial multiplexing. Results show that spatial diversity MIMO VLC improves\nerror performance at the cost of spectral efficiency that spatial multiplexing\nshould enhance. We propose the adaptive MIMO solution that both modulation\nschema and MIMO schema are dynamically adapted to the changing channel\nconditions for enhancing the error performance and spectral efficiency. The\naverage error-free spectral efficiency of adaptive 2x2 MIMO VLC achieved 12\nb/s/Hz over 2 meters indoor dynamic transmission.\n",
"title": "Adaptive Real-Time Software Defined MIMO Visible Light Communications using Spatial Multiplexing and Spatial Diversity"
}
| null | null | null | null | true | null |
17262
| null |
Default
| null | null |
null |
{
"abstract": " The continuous dynamical system approach to deep learning is explored in\norder to devise alternative frameworks for training algorithms. Training is\nrecast as a control problem and this allows us to formulate necessary\noptimality conditions in continuous time using the Pontryagin's maximum\nprinciple (PMP). A modification of the method of successive approximations is\nthen used to solve the PMP, giving rise to an alternative training algorithm\nfor deep learning. This approach has the advantage that rigorous error\nestimates and convergence results can be established. We also show that it may\navoid some pitfalls of gradient-based methods, such as slow convergence on flat\nlandscapes near saddle points. Furthermore, we demonstrate that it obtains\nfavorable initial convergence rate per-iteration, provided Hamiltonian\nmaximization can be efficiently carried out - a step which is still in need of\nimprovement. Overall, the approach opens up new avenues to attack problems\nassociated with deep learning, such as trapping in slow manifolds and\ninapplicability of gradient-based methods for discrete trainable variables.\n",
"title": "Maximum Principle Based Algorithms for Deep Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17263
| null |
Validated
| null | null |
null |
{
"abstract": " With the popularity of Linked Open Data (LOD) and the associated rise in\nfreely accessible knowledge that can be accessed via LOD, exploiting LOD for\nrecommender systems has been widely studied based on various approaches such as\ngraph-based or using different machine learning models with LOD-enabled\nfeatures. Many of the previous approaches require construction of an additional\ngraph to run graph-based algorithms or to extract path-based features by\ncombining user- item interactions (e.g., likes, dislikes) and background\nknowledge from LOD. In this paper, we investigate Factorization Machines (FMs)\nbased on particularly lightweight LOD-enabled features which can be directly\nobtained via a public SPARQL Endpoint without any additional effort to\nconstruct a graph. Firstly, we aim to study whether using FM with these\nlightweight LOD-enabled features can provide competitive performance compared\nto a learning-to-rank approach leveraging LOD as well as other well-established\napproaches such as kNN-item and BPRMF. Secondly, we are interested in finding\nout to what extent each set of LOD-enabled features contributes to the\nrecommendation performance. Experimental evaluation on a standard dataset shows\nthat our proposed approach using FM with lightweight LOD-enabled features\nprovides the best performance compared to other approaches in terms of five\nevaluation metrics. In addition, the study of the recommendation performance\nbased on different sets of LOD-enabled features indicate that property-object\nlists and PageRank scores of items are useful for improving the performance,\nand can provide the best performance through using them together for FM. We\nobserve that subject-property lists of items does not contribute to the\nrecommendation performance but rather decreases the performance.\n",
"title": "Factorization Machines Leveraging Lightweight Linked Open Data-enabled Features for Top-N Recommendations"
}
| null | null | null | null | true | null |
17264
| null |
Default
| null | null |
null |
{
"abstract": " Wave motion in two- and three-dimensional periodic lattices of beam members\nsupporting longitudinal and flexural waves is considered. An analytic method\nfor solving the Bloch wave spectrum is developed, characterized by a\ngeneralized eigenvalue equation obtained by enforcing the Floquet condition.\nThe dynamic stiffness matrix is shown to be explicitly Hermitian and to admit\npositive eigenvalues. Lattices with hexagonal, rectangular, tetrahedral and\ncubic unit cells are analyzed. The semi-analytical method can be asymptotically\nexpanded for low frequency yielding explicit forms for the Christoffel matrix\ndescribing wave motion in the quasistatic limit.\n",
"title": "Wave propagation and homogenization in 2D and 3D lattices: a semi-analytical approach"
}
| null | null | null | null | true | null |
17265
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we formulate an analogue of Waring's problem for an algebraic\ngroup $G$. At the field level we consider a morphism of varieties $f\\colon\n\\mathbb{A}^1\\to G$ and ask whether every element of $G(K)$ is the product of a\nbounded number of elements $f(\\mathbb{A}^1(K)) = f(K)$. We give an affirmative\nanswer when $G$ is unipotent and $K$ is a characteristic zero field which is\nnot formally real.\nThe idea is the same at the integral level, except one must work with\nschemes, and the question is whether every element in a finite index subgroup\nof $G(\\mathcal{O})$ can be written as a product of a bounded number of elements\nof $f(\\mathcal{O})$. We prove this is the case when $G$ is unipotent and\n$\\mathcal{O}$ is the ring of integers of a totally imaginary number field.\n",
"title": "Waring's problem for unipotent algebraic groups"
}
| null | null | null | null | true | null |
17266
| null |
Default
| null | null |
null |
{
"abstract": " Many real-world multilayer systems such as critical infrastructure are\ninterdependent and embedded in space with links of a characteristic length.\nThey are also vulnerable to localized attacks or failures, such as terrorist\nattacks or natural catastrophes, which affect all nodes within a given radius.\nHere we study the effects of localized attacks on spatial multiplex networks of\ntwo layers. We find a metastable region where a localized attack larger than a\ncritical size induces a nucleation transition as a cascade of failures spreads\nthroughout the system, leading to its collapse. We develop a theory to predict\nthe critical attack size and find that it exhibits novel scaling behavior. We\nfurther find that localized attacks in these multiplex systems can induce a\npreviously unobserved combination of random and spatial cascades. Our results\ndemonstrate important vulnerabilities in real-world interdependent networks and\nshow new theoretical features of spatial networks.\n",
"title": "Spreading of localized attacks in spatial multiplex networks"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
17267
| null |
Validated
| null | null |
null |
{
"abstract": " The reconstruction of sparse signals requires the solution of an\n$\\ell_0$-norm minimization problem in Compressed Sensing. Previous research has\nfocused on the investigation of a single candidate to identify the support\n(index of nonzero elements) of a sparse signal. To ensure that the optimal\ncandidate can be obtained in each iteration, we propose here an iterative\ngreedy reconstruction algorithm (GSRA). First, the intersection of the support\nsets estimated by the Orthogonal Matching Pursuit (OMP) and Subspace Pursuit\n(SP) is set as the initial support set. Then, a hope-tree is built to expand\nthe set. Finally, a developed decreasing subspace pursuit method is used to\nrectify the candidate set. Detailed simulation results demonstrate that GSRA is\nmore accurate than other typical methods in recovering Gaussian signals, 0--1\nsparse signals, and synthetic signals.\n",
"title": "Greedy Sparse Signal Reconstruction Using Matching Pursuit Based on Hope-tree"
}
| null | null | null | null | true | null |
17268
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose a fault detection and isolation based attack-aware\nmulti-sensor integration algorithm for the detection of cyberattacks in\nautonomous vehicle navigation systems. The proposed algorithm uses an extended\nKalman filter to construct robust residuals in the presence of noise, and then\nuses a parametric statistical tool to identify cyberattacks. The parametric\nstatistical tool is based on the residuals constructed by the measurement\nhistory rather than one measurement at a time in the properties of\ndiscrete-time signals and dynamic systems. This approach allows the proposed\nmulti-sensor integration algorithm to provide quick detection and low false\nalarm rates for applications in dynamic systems. An example of INS/GNSS\nintegration of autonomous navigation systems is presented to validate the\nproposed algorithm by using a software-in-the-loop simulation.\n",
"title": "Attack-Aware Multi-Sensor Integration Algorithm for Autonomous Vehicle Navigation Systems"
}
| null | null | null | null | true | null |
17269
| null |
Default
| null | null |
null |
{
"abstract": " We study numerically a Constantin-Lax-Majda-De Gregorio model generalized by\nOkamoto, Sakajo and Wunsch, which is a model of fluid turbulence in one\ndimension with an inviscid conservation law. In the presence of the viscosity\nand two types of the large-scale forcings, we show that turbulent cascade of\nthe inviscid invariant, which is not limited to quadratic quantity, occurs and\nthat properties of this model's turbulent state are related to singularity of\nthe inviscid case by adopting standard tools of analyzing fluid turbulence.\n",
"title": "Turbulence, cascade and singularity in a generalization of the Constantin-Lax-Majda equation"
}
| null | null | null | null | true | null |
17270
| null |
Default
| null | null |
null |
{
"abstract": " We consider the fitting of heavy tailed data and distribution with a special\nattention to distributions with a non--standard shape in the \"body\" of the\ndistribution. To this end we consider a dense class of heavy tailed\ndistributions introduced recently, employing an EM algorithm for the the\nmaximum likelihood estimates of its parameters. We present methods for fitting\nto observed data, histograms, censored data, as well as to theoretical\ndistributions. Numerical examples are provided with simulated data and a\nbenchmark reinsurance dataset. We empirically demonstrate that our model can\nprovide excellent fits to heavy--tailed data/distributions with minimal\nassumptions\n",
"title": "Fitting phase--type scale mixtures to heavy--tailed data and distributions"
}
| null | null | null | null | true | null |
17271
| null |
Default
| null | null |
null |
{
"abstract": " This paper introduces Deep Incremental Boosting, a new technique derived from\nAdaBoost, specifically adapted to work with Deep Learning methods, that reduces\nthe required training time and improves generalisation. We draw inspiration\nfrom Transfer of Learning approaches to reduce the start-up time to training\neach incremental Ensemble member. We show a set of experiments that outlines\nsome preliminary results on some common Deep Learning datasets and discuss the\npotential improvements Deep Incremental Boosting brings to traditional Ensemble\nmethods in Deep Learning.\n",
"title": "Deep Incremental Boosting"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17272
| null |
Validated
| null | null |
null |
{
"abstract": " We consider linear structural equation models that are associated with mixed\ngraphs. The structural equations in these models only involve observed\nvariables, but their idiosyncratic error terms are allowed to be correlated and\nnon-Gaussian. We propose empirical likelihood (EL) procedures for inference,\nand suggest several modifications, including a profile likelihood, in order to\nimprove tractability and performance of the resulting methods. Through\nsimulations, we show that when the error distributions are non-Gaussian, the\nuse of EL and the proposed modifications may increase statistical efficiency\nand improve assessment of significance.\n",
"title": "Empirical Likelihood for Linear Structural Equation Models with Dependent Errors"
}
| null | null | null | null | true | null |
17273
| null |
Default
| null | null |
null |
{
"abstract": " We show how solutions to a large class of partial differential equations with\nnonlocal Riccati-type nonlinearities can be generated from the corresponding\nlinearized equations, from arbitrary initial data. It is well known that\nevolutionary matrix Riccati equations can be generated by projecting linear\nevolutionary flows on a Stiefel manifold onto a coordinate chart of the\nunderlying Grassmann manifold. Our method relies on extending this idea to the\ninfinite dimensional case. The key is an integral equation analogous to the\nMarchenko equation in integrable systems, that represents the coodinate chart\nmap. We show explicitly how to generate such solutions to scalar partial\ndifferential equations of arbitrary order with nonlocal quadratic\nnonlinearities using our approach. We provide numerical simulations that\ndemonstrate the generation of solutions to\nFisher--Kolmogorov--Petrovskii--Piskunov equations with nonlocal\nnonlinearities. We also indicate how the method might extend to more general\nclasses of nonlinear partial differential systems.\n",
"title": "Grassmannian flows and applications to nonlinear partial differential equations"
}
| null | null |
[
"Physics",
"Mathematics"
] | null | true | null |
17274
| null |
Validated
| null | null |
null |
{
"abstract": " In 1934, Reinhardt conjectured that the shape of the centrally symmetric\nconvex body in the plane whose densest lattice packing has the smallest density\nis a smoothed octagon. This conjecture is still open. We formulate the\nReinhardt Conjecture as a problem in optimal control theory. The smoothed\noctagon is a Pontryagin extremal trajectory with bang-bang control. More\ngenerally, the smoothed regular $6k+2$-gon is a Pontryagin extremal with\nbang-bang control. The smoothed octagon is a strict (micro) local minimum to\nthe optimal control problem. The optimal solution to the Reinhardt problem is a\ntrajectory without singular arcs. The extremal trajectories that do not meet\nthe singular locus have bang-bang controls with finitely many switching times.\nFinally, we reduce the Reinhardt problem to an optimization problem on a\nfive-dimensional manifold. (Each point on the manifold is an initial condition\nfor a potential Pontryagin extremal lifted trajectory.) We suggest that the\nReinhardt conjecture might eventually be fully resolved through optimal control\ntheory. Some proofs are computer-assisted using a computer algebra system.\n",
"title": "The Reinhardt Conjecture as an Optimal Control Problem"
}
| null | null | null | null | true | null |
17275
| null |
Default
| null | null |
null |
{
"abstract": " We study the heating mechanisms and Ly{\\alpha} escape fractions of 35\nLy{\\alpha} blobs (LABs) at z = 3.1 in the SSA22 field. Dust continuum sources\nhave been identified in 11 of the 35 LABs, all with star formation rates (SFRs)\nabove 100 Msun/yr. Likely radio counterparts are detected in 9 out of 29\ninvestigated LABs. The detection of submm dust emission is more linked to the\nphysical size of the Ly{\\alpha} emission than to the Ly{\\alpha} luminosities of\nthe LABs. A radio excess in the submm/radio detected LABs is common, hinting at\nthe presence of active galactic nuclei. Most radio sources without X-ray\ncounterparts are located at the centers of the LABs. However, all X-ray\ncounterparts avoid the central regions. This may be explained by absorption due\nto exceptionally large column densities along the line-of-sight or by LAB\nmorphologies, which are highly orientation dependent. The median Ly{\\alpha}\nescape fraction is about 3\\% among the submm-detected LABs, which is lower than\na lower limit of 11\\% for the submm-undetected LABs. We suspect that the large\ndifference is due to the high dust attenuation supported by the large SFRs, the\ndense large-scale environment as well as large uncertainties in the extinction\ncorrections required to apply when interpreting optical data.\n",
"title": "Deep submillimeter and radio observations in the SSA22 field. I. Powering sources and Lyα escape fraction of Lyα blobs"
}
| null | null | null | null | true | null |
17276
| null |
Default
| null | null |
null |
{
"abstract": " In this chapter we explain briefly the fundamentals of the interactive scores\nformalism. Then we develop a solution for implementing the ECO machine by\nmixing petri nets and constraints propagation. We also present another solution\nfor implementing the ECO machine using concurrent constraint programming.\nFinally, we present an extension of interactive score with conditional\nbranching.\n",
"title": "Modeling temporal constraints for a system of interactive scores"
}
| null | null | null | null | true | null |
17277
| null |
Default
| null | null |
null |
{
"abstract": " The electronic structure of ThRu2Si2 was studied by angle-resolved\nphotoelectron spectroscopy (ARPES) with incident photon energies of hn=655-745\neV. Detailed band structure and the three-dimensional shapes of Fermi surfaces\nwere derived experimentally, and their characteristic features were mostly\nexplained by means of band structure calculations based on the density\nfunctional theory. Comparison of the experimental ARPES spectra of ThRu2Si2\nwith those of URu2Si2 shows that they have considerably different spectral\nprofiles particularly in the energy range of 1 eV from the Fermi level,\nsuggesting that U 5f states are substantially hybridized in these bands. The\nrelationship between the ARPES spectra of URu2Si2 and ThRu2Si2 is very\ndifferent from the one between the ARPES spectra of CeRu2Si2 and LaRu2Si2,\nwhere the intrinsic difference in their spectra is limited only in the very\nvicinity of the Fermi energy. The present result suggests that the U 5f\nelectrons in URu2Si2 have strong hybridization with ligand states and have an\nessentially itinerant character.\n",
"title": "Electronic structure of ThRu2Si2 studied by angle-resolved photoelectron spectroscopy: Elucidating the contribution of U 5f states in URu2Si2"
}
| null | null |
[
"Physics"
] | null | true | null |
17278
| null |
Validated
| null | null |
null |
{
"abstract": " Factorable surfaces, i.e. graphs associated with the product of two functions\nof one variable, constitute a wide class of surfaces. Such surfaces in the\npseudo-Galilean space with zero Gaussian and mean curvature were obtained in\n[1]. In this study, we provide new classification results relating to the\nfactorable surfaces with non-zero Gaussian and mean curvature.\n",
"title": "Non-zero constant curvature factorable surfaces in pseudo-Galilean space"
}
| null | null | null | null | true | null |
17279
| null |
Default
| null | null |
null |
{
"abstract": " The paper presents two results. First it is shown how the discrete potential\nmodified KdV equation and its Lax pairs in matrix form arise from the\nHirota-Miwa equation by a 2-periodic reduction. Then Darboux transformations\nand binary Darboux transformations are derived for the discrete potential\nmodified KdV equation and it is shown how these may be used to construct exact\nsolutions.\n",
"title": "Darboux and Binary Darboux Transformations for Discrete Integrable Systems. II. Discrete Potential mKdV Equation"
}
| null | null | null | null | true | null |
17280
| null |
Default
| null | null |
null |
{
"abstract": " Sequential change-point detection when the distribution parameters are\nunknown is a fundamental problem in statistics and machine learning. When the\npost-change parameters are unknown, we consider a set of detection procedures\nbased on sequential likelihood ratios with non-anticipating estimators\nconstructed using online convex optimization algorithms such as online mirror\ndescent, which provides a more versatile approach to tackle complex situations\nwhere recursive maximum likelihood estimators cannot be found. When the\nunderlying distributions belong to a exponential family and the estimators\nsatisfy the logarithm regret property, we show that this approach is nearly\nsecond-order asymptotically optimal. This means that the upper bound for the\nfalse alarm rate of the algorithm (measured by the average-run-length) meets\nthe lower bound asymptotically up to a log-log factor when the threshold tends\nto infinity. Our proof is achieved by making a connection between sequential\nchange-point and online convex optimization and leveraging the logarithmic\nregret bound property of online mirror descent algorithm. Numerical and real\ndata examples validate our theory.\n",
"title": "Nearly second-order asymptotic optimality of sequential change-point detection with one-sample updates"
}
| null | null | null | null | true | null |
17281
| null |
Default
| null | null |
null |
{
"abstract": " We give algorithms to construct the Néron Desingularization and the easy\ncase from \\cite{KK} of the General Néron Desingularization.\n",
"title": "Algorithms in the classical Néron Desingularization"
}
| null | null | null | null | true | null |
17282
| null |
Default
| null | null |
null |
{
"abstract": " In recent years, deep learning has made tremendous progress in a number of\nfields that were previously out of reach for artificial intelligence. The\nsuccesses in these problems has led researchers to consider the possibilities\nfor intelligent systems to tackle a problem that humans have only recently\nthemselves considered: program synthesis. This challenge is unlike others such\nas object recognition and speech translation, since its abstract nature and\ndemand for rigor make it difficult even for human minds to attempt. While it is\nstill far from being solved or even competitive with most existing methods,\nneural program synthesis is a rapidly growing discipline which holds great\npromise if completely realized. In this paper, we start with exploring the\nproblem statement and challenges of program synthesis. Then, we examine the\nfascinating evolution of program induction models, along with how they have\nsucceeded, failed and been reimagined since. Finally, we conclude with a\ncontrastive look at program synthesis and future research recommendations for\nthe field.\n",
"title": "Recent Advances in Neural Program Synthesis"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17283
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the problem of training generative models with deep neural\nnetworks as generators, i.e. to map latent codes to data points. Whereas the\ndominant paradigm combines simple priors over codes with complex deterministic\nmodels, we propose instead to use more flexible code distributions. These\ndistributions are estimated non-parametrically by reversing the generator map\nduring training. The benefits include: more powerful generative models, better\nmodeling of latent structure and explicit control of the degree of\ngeneralization.\n",
"title": "Generator Reversal"
}
| null | null | null | null | true | null |
17284
| null |
Default
| null | null |
null |
{
"abstract": " Ontology-based query answering (OBQA) asks whether a Boolean conjunctive\nquery is satisfied by all models of a logical theory consisting of a relational\ndatabase paired with an ontology. The introduction of existential rules (i.e.,\nDatalog rules extended with existential quantifiers in rule-heads) as a means\nto specify the ontology gave birth to Datalog+/-, a framework that has received\nincreasing attention in the last decade, with focus also on decidability and\nfinite controllability to support effective reasoning. Five basic decidable\nfragments have been singled out: linear, weakly-acyclic, guarded, sticky, and\nshy. Moreover, for all these fragments, except shy, the important property of\nfinite controllability has been proved, ensuring that a query is satisfied by\nall models of the theory iff it is satisfied by all its finite models. In this\npaper we complete the picture by demonstrating that finite controllability of\nOBQA holds also for shy ontologies, and it therefore applies to all basic\ndecidable Datalog+/- classes. To make the demonstration, we devise a general\ntechnique to facilitate the process of (dis)proving finite controllability of\nan arbitrary ontological fragment. This paper is under consideration for\nacceptance in TPLP.\n",
"title": "Finite model reasoning over existential rules"
}
| null | null | null | null | true | null |
17285
| null |
Default
| null | null |
null |
{
"abstract": " Despite their popularity, the practical performance of asynchronous\nstochastic gradient descent methods (ASGD) for solving large scale machine\nlearning problems are not as good as theoretical results indicate. We adopt and\nanalyze a synchronous K-step averaging stochastic gradient descent algorithm\nwhich we call K-AVG. We establish the convergence results of K-AVG for\nnonconvex objectives and explain why the K-step delay is necessary and leads to\nbetter performance than traditional parallel stochastic gradient descent which\nis a special case of K-AVG with $K=1$. We also show that K-AVG scales better\nthan ASGD. Another advantage of K-AVG over ASGD is that it allows larger\nstepsizes. On a cluster of $128$ GPUs, K-AVG is faster than ASGD\nimplementations and achieves better accuracies and faster convergence for\n\\cifar dataset.\n",
"title": "On the convergence properties of a $K$-step averaging stochastic gradient descent algorithm for nonconvex optimization"
}
| null | null | null | null | true | null |
17286
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study a new learning paradigm for Neural Machine\nTranslation (NMT). Instead of maximizing the likelihood of the human\ntranslation as in previous works, we minimize the distinction between human\ntranslation and the translation given by an NMT model. To achieve this goal,\ninspired by the recent success of generative adversarial networks (GANs), we\nemploy an adversarial training architecture and name it as Adversarial-NMT. In\nAdversarial-NMT, the training of the NMT model is assisted by an adversary,\nwhich is an elaborately designed Convolutional Neural Network (CNN). The goal\nof the adversary is to differentiate the translation result generated by the\nNMT model from that by human. The goal of the NMT model is to produce high\nquality translations so as to cheat the adversary. A policy gradient method is\nleveraged to co-train the NMT model and the adversary. Experimental results on\nEnglish$\\rightarrow$French and German$\\rightarrow$English translation tasks\nshow that Adversarial-NMT can achieve significantly better translation quality\nthan several strong baselines.\n",
"title": "Adversarial Neural Machine Translation"
}
| null | null | null | null | true | null |
17287
| null |
Default
| null | null |
null |
{
"abstract": " We determine which amalgamated products of surface groups identified over\nmultiples of simple closed curves are not fundamental groups of 3-manifolds. We\nprove each surface amalgam considered is virtually the fundamental group of a\n3-manifold. We prove that each such surface group amalgam is abstractly\ncommensurable to a right-angled Coxeter group from a related family. In an\nappendix, we determine the quasi-isometry classes among these surface amalgams\nand their related right-angled Coxeter groups.\n",
"title": "Surface group amalgams that (don't) act on 3-manifolds"
}
| null | null | null | null | true | null |
17288
| null |
Default
| null | null |
null |
{
"abstract": " Understanding shading effects in images is critical for a variety of vision\nand graphics problems, including intrinsic image decomposition, shadow removal,\nimage relighting, and inverse rendering. As is the case with other vision\ntasks, machine learning is a promising approach to understanding shading - but\nthere is little ground truth shading data available for real-world images. We\nintroduce Shading Annotations in the Wild (SAW), a new large-scale, public\ndataset of shading annotations in indoor scenes, comprised of multiple forms of\nshading judgments obtained via crowdsourcing, along with shading annotations\nautomatically generated from RGB-D imagery. We use this data to train a\nconvolutional neural network to predict per-pixel shading information in an\nimage. We demonstrate the value of our data and network in an application to\nintrinsic images, where we can reduce decomposition artifacts produced by\nexisting algorithms. Our database is available at\nthis http URL.\n",
"title": "Shading Annotations in the Wild"
}
| null | null | null | null | true | null |
17289
| null |
Default
| null | null |
null |
{
"abstract": " Let $S$ be the power series ring or the polynomial ring over a field $K$ in\nthe variables $x_1,\\ldots,x_n$, and let $R=S/I$, where $I$ is proper ideal\nwhich we assume to be graded if $S$ is the polynomial ring. We give an explicit\ndescription of the cycles of the Koszul complex whose homology classes generate\nthe Koszul homology of $R=S/I$ with respect to $x_1,\\ldots,x_n$. The\ndescription is given in terms of the data of the free $S$-resolution of $R$.\nThe result is used to determine classes of Golod ideals, among them proper\nordinary powers and proper symbolic powers of monomial ideals. Our theory is\nalso applied to stretched local rings.\n",
"title": "Koszul cycles and Golod rings"
}
| null | null | null | null | true | null |
17290
| null |
Default
| null | null |
null |
{
"abstract": " Generative adversarial networks (GANs) are innovative techniques for learning\ngenerative models of complex data distributions from samples. Despite\nremarkable recent improvements in generating realistic images, one of their\nmajor shortcomings is the fact that in practice, they tend to produce samples\nwith little diversity, even when trained on diverse datasets. This phenomenon,\nknown as mode collapse, has been the main focus of several recent advances in\nGANs. Yet there is little understanding of why mode collapse happens and why\nexisting approaches are able to mitigate mode collapse. We propose a principled\napproach to handling mode collapse, which we call packing. The main idea is to\nmodify the discriminator to make decisions based on multiple samples from the\nsame class, either real or artificially generated. We borrow analysis tools\nfrom binary hypothesis testing---in particular the seminal result of Blackwell\n[Bla53]---to prove a fundamental connection between packing and mode collapse.\nWe show that packing naturally penalizes generators with mode collapse, thereby\nfavoring generator distributions with less mode collapse during the training\nprocess. Numerical experiments on benchmark datasets suggests that packing\nprovides significant improvements in practice as well.\n",
"title": "PacGAN: The power of two samples in generative adversarial networks"
}
| null | null | null | null | true | null |
17291
| null |
Default
| null | null |
null |
{
"abstract": " Causal mediation analysis aims to estimate the natural direct and indirect\neffects under clearly specified assumptions. Traditional mediation analysis\nbased on Ordinary Least Squares (OLS) relies on the absence of unmeasured\ncauses of the putative mediator and outcome. When this assumption cannot be\njustified, Instrumental Variables (IV) estimators can be used in order to\nproduce an asymptotically unbiased estimator of the mediator-outcome link.\nHowever, provided that valid instruments exist, bias removal comes at the cost\nof variance inflation for standard IV procedures such as Two-Stage Least\nSquares (TSLS). A Semi-Parametric Stein-Like (SPSL) estimator has been proposed\nin the literature that strikes a natural trade-off between the unbiasedness of\nthe TSLS procedure and the relatively small variance of the OLS estimator.\nMoreover, the SPSL has the advantage that its shrinkage parameter can be\ndirectly estimated from the data. In this paper, we demonstrate how this\nStein-like estimator can be implemented in the context of the estimation of\nnatural direct and natural indirect effects of treatments in randomized\ncontrolled trials. The performance of the competing methods is studied in a\nsimulation study, in which both the strength of hidden confounding and the\nstrength of the instruments are independently varied. These considerations are\nmotivated by a trial in mental health evaluating the impact of a primary\ncare-based intervention to reduce depression in the elderly.\n",
"title": "Stein-like Estimators for Causal Mediation Analysis in Randomized Trials"
}
| null | null | null | null | true | null |
17292
| null |
Default
| null | null |
null |
{
"abstract": " In this work, a novel subspace-based method for blind identification of\nmultichannel finite impulse response (FIR) systems is presented. Here, we\nexploit directly the impeded Toeplitz channel structure in the signal linear\nmodel to build a quadratic form whose minimization leads to the desired channel\nestimation up to a scalar factor. This method can be extended to estimate any\npredefined linear structure, e.g. Hankel, that is usually encountered in linear\nsystems. Simulation findings are provided to highlight the appealing advantages\nof the new structure-based subspace (SSS) method over the standard subspace\n(SS) method in certain adverse identification scenarii.\n",
"title": "Structure-Based Subspace Method for Multi-Channel Blind System Identification"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17293
| null |
Validated
| null | null |
null |
{
"abstract": " We extend a previously introduced semi-analytical representation of a\ndecomposition of CA dynamics in arbitrary dimensions and neighborhood schemes\nvia the use of certain universal maps in which CA rule vectors are derivable\nfrom the equivalent of superpotentials. The results justify the search for\nalternative analog models of computation and their possible physical\nconnections.\n",
"title": "On Certain Analytical Representations of Cellular Automata"
}
| null | null | null | null | true | null |
17294
| null |
Default
| null | null |
null |
{
"abstract": " In this article we study the existence and strong consistency of GEE\nestimators, when the generalized estimating functions are martingales with\nrandom coefficients. Furthermore, we characterize estimating functions which\nare asymptotically optimal.\n",
"title": "Strong consistency and optimality for generalized estimating equations with stochastic covariates"
}
| null | null | null | null | true | null |
17295
| null |
Default
| null | null |
null |
{
"abstract": " We report on the selective fabrication of high-quality Sr$_2$IrO$_4$ and\nSrIrO$_3$ epitaxial thin films from a single polycrystalline Sr$_2$IrO$_4$\ntarget by pulsed laser deposition. Using a combination of X-ray diffraction and\nphotoemission spectroscopy characterizations, we discover that within a\nrelatively narrow range of substrate temperature, the oxygen partial pressure\nplays a critical role in the cation stoichiometric ratio of the films, and\ntriggers the stabilization of different Ruddlesden-Popper (RP) phases. Resonant\nX-ray absorption spectroscopy measurements taken at the Ir $L$-edge and the O\n$K$-edge demonstrate the presence of strong spin-orbit coupling, and reveal the\nelectronic and orbital structures of both compounds. These results suggest that\nin addition to the conventional thermodynamics consideration, higher members of\nthe Sr$_{n+1}$Ir$_n$O$_{3n+1}$ series can possibly be achieved by kinetic\ncontrol away from the thermodynamic limit. These findings offer a new approach\nto the synthesis of ultra-thin films of the RP series of iridates and can be\nextended to other complex oxides with layered structure.\n",
"title": "Synthesis and electronic properties of Ruddlesden-Popper strontium iridate epitaxial thin films stabilized by control of growth kinetics"
}
| null | null | null | null | true | null |
17296
| null |
Default
| null | null |
null |
{
"abstract": " In this note, we prove an ${L^{\\frac{n}{2}}}$-energy gap result for\nYang-Mills connections on a principal $G$-bundle over a compact manifold\nwithout using Lojasiewicz-Simon gradient inequality (arXiv:1502.00668).\n",
"title": "A proof on energy gap for Yang-Mills connection"
}
| null | null | null | null | true | null |
17297
| null |
Default
| null | null |
null |
{
"abstract": " Pomsets are a model of concurrent computations introduced by Pratt. They can\nprovide a syntax-oblivious description of semantics of coordination models\nbased on asynchronous message-passing, such as Message Sequence Charts (MSCs).\nIn this paper, we study conditions that ensure a specification expressed as a\nset of pomsets can be faithfully realised via communicating automata. Our main\ncontributions are (i) the definition of a realisability condition accounting\nfor termination soundness, (ii) conditions for global specifications with\n\"multi-threaded\" participants, and (iii) the definition of realisability\nconditions that can be decided directly over pomsets. A positive by-product of\nour approach is the efficiency gain in the verification of the realisability\nconditions obtained when restricting to specific classes of choreographies\ncharacterisable in term of behavioural types.\n",
"title": "Realisability of Pomsets via Communicating Automata"
}
| null | null | null | null | true | null |
17298
| null |
Default
| null | null |
null |
{
"abstract": " The ecological invasion problem in which a weaker exotic species invades an\necosystem inhabited by two strongly competing native species is modelled by a\nthree-species competition-diffusion system. It is known that for a certain\nrange of parameter values competitor-mediated coexistence occurs and complex\nspatio-temporal patterns are observed in two spatial dimensions. In this paper\nwe uncover the mechanism which generates such patterns. Under some assumptions\non the parameters the three-species competition-diffusion system admits two\nplanarly stable travelling waves. Their interaction in one spatial dimension\nmay result in either reflection or merging into a single homoclinic wave,\ndepending on the strength of the invading species. This transition can be\nunderstood by studying the bifurcation structure of the homoclinic wave. In\nparticular, a time-periodic homoclinic wave (breathing wave) is born from a\nHopf bifurcation and its unstable branch acts as a separator between the\nreflection and merging regimes. The same transition occurs in two spatial\ndimensions: the stable regular spiral associated to the homoclinic wave\ndestabilizes, giving rise first to an oscillating breathing spiral and then\nbreaking up producing a dynamic pattern characterized by many spiral cores. We\nfind that these complex patterns are generated by the interaction of two\nplanarly stable travelling waves, in contrast with many other well known cases\nof pattern formation where planar instability plays a central role.\n",
"title": "Complex pattern formation driven by the interaction of stable fronts in a competition-diffusion system"
}
| null | null | null | null | true | null |
17299
| null |
Default
| null | null |
null |
{
"abstract": " Nonlocality is a key feature of many physical systems since it prevents a\ncatastrophic collapse and a symmetry-breaking azimuthal instability of intense\nwave beams in a bulk self-focusing nonlinear media. This opens up an intriguing\nperspective for stabilization of complex topological structures such as\nhigher-order solitons, vortex rings and vortex ring-on-line complexes. Using\ndirect numerical simulations, we find a class of cylindrically-symmetric $n$-th\norder spatial solitons having the intensity distribution with a central bright\nspot surrounded by $n$ bright rings of varying size. We investigate dynamical\nproperties of these higher-order solitons in a media with thermal nonlocal\nnonlinear response. We show theoretically that a vortex complex of vortex ring\nand vortex line, carrying two independent winding numbers, can be created by\nperturbation of the stable optical vortex soliton in nonlocal nonlinear media.\n",
"title": "Solitons with rings and vortex rings on solitons in nonlocal nonlinear media"
}
| null | null |
[
"Physics"
] | null | true | null |
17300
| null |
Validated
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.