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": " Coherent control of the resonant response in spatially extended\noptomechanical structures is complicated by the fact that the optical drive is\naffected by the back-action from the generated phonons. Here we report a new\napproach to coherent control based on stimulated Raman-like scattering, in\nwhich the optical pressure can remain unaffected by the induced vibrations even\nin the regime of strong optomechanical interactions. We demonstrate\nexperimentally coherent control of flexural vibrations simultaneously along the\nwhole length of a dual-nanoweb fiber, by imprinting steps in the relative phase\nbetween the components of a two-frequency pump signal,the beat frequency being\nchosen to match a flexural resonance. Furthermore, sequential switching of the\nrelative phase at time intervals shorter than the lifetime of the vibrations\nreduces their amplitude to a constant value that is fully adjustable by tuning\nthe phase-modulation depth and switching rate. The results may trigger new\ndevelopments in silicon photonics, since such coherent control uniquely\ndecouples the amplitude of optomechanical oscillations from power-dependent\nthermal effects and nonlinear optical loss.\n",
"title": "Coherent control of flexural vibrations in dual-nanoweb fibers using phase-modulated two-frequency light"
}
| null | null | null | null | true | null |
17901
| null |
Default
| null | null |
null |
{
"abstract": " Recently we have demonstrated the presence of spin-orbit toque in FeMn/Pt\nmultilayers which, in combination with the anisotropy field, is able to rotate\nits magnetization consecutively from 0o to 360o without any external field.\nHere, we report on an investigation of static and dynamic magnetic properties\nof FeMn/Pt multilayers using combined techniques of magnetometry, ferromagnetic\nresonance, inverse spin Hall effect and spin Hall magnetoresistance\nmeasurements. The FeMn/Pt multilayer was found to exhibit ferromagnetic\nproperties, and its temperature dependence of saturation magnetization can be\nfitted well using a phenomenological model by including a finite distribution\nin Curie temperature due to subtle thickness variations across the multilayer\nsamples. The non-uniformity in static magnetic properties is also manifested in\nthe ferromagnetic resonance spectra, which typically exhibit a broad resonance\npeak. A damping parameter of around 0.106 is derived from the frequency\ndependence of ferromagnetic resonance linewidth, which is comparable to the\nreported values for other types of Pt-based multilayers. Clear inverse spin\nHall signals and spin Hall magnetoresistance have been observed in all samples\nbelow the Curie temperature, which corroborate the strong spin-orbit torque\neffect observed previously.\n",
"title": "Static and Dynamic Magnetic Properties of FeMn/Pt Multilayers"
}
| null | null | null | null | true | null |
17902
| null |
Default
| null | null |
null |
{
"abstract": " We study a relationship between the ultraproduct of a crossed product von\nNeumann algebra and the crossed product of an ultraproduct von Neumann algebra.\nAs an application, the continuous core of an ultraproduct von Neumann algebra\nis described.\n",
"title": "Ultraproducts of crossed product von Neumann algebras"
}
| null | null |
[
"Mathematics"
] | null | true | null |
17903
| null |
Validated
| null | null |
null |
{
"abstract": " A fundamental issue for statistical classification models in a streaming\nenvironment is that the joint distribution between predictor and response\nvariables changes over time (a phenomenon also known as concept drifts), such\nthat their classification performance deteriorates dramatically. In this paper,\nwe first present a hierarchical hypothesis testing (HHT) framework that can\ndetect and also adapt to various concept drift types (e.g., recurrent or\nirregular, gradual or abrupt), even in the presence of imbalanced data labels.\nA novel concept drift detector, namely Hierarchical Linear Four Rates (HLFR),\nis implemented under the HHT framework thereafter. By substituting a\nwidely-acknowledged retraining scheme with an adaptive training strategy, we\nfurther demonstrate that the concept drift adaptation capability of HLFR can be\nsignificantly boosted. The theoretical analysis on the Type-I and Type-II\nerrors of HLFR is also performed. Experiments on both simulated and real-world\ndatasets illustrate that our methods outperform state-of-the-art methods in\nterms of detection precision, detection delay as well as the adaptability\nacross different concept drift types.\n",
"title": "Concept Drift Detection and Adaptation with Hierarchical Hypothesis Testing"
}
| null | null | null | null | true | null |
17904
| null |
Default
| null | null |
null |
{
"abstract": " The problem of machine learning with missing values is common in many areas.\nA simple approach is to first construct a dataset without missing values simply\nby discarding instances with missing entries or by imputing a fixed value for\neach missing entry, and then train a prediction model with the new dataset. A\ndrawback of this naive approach is that the uncertainty in the missing entries\nis not properly incorporated in the prediction. In order to evaluate prediction\nuncertainty, the multiple imputation (MI) approach has been studied, but the\nperformance of MI is sensitive to the choice of the probabilistic model of the\ntrue values in the missing entries, and the computational cost of MI is high\nbecause multiple models must be trained. In this paper, we propose an\nalternative approach called the Interval-based Prediction Uncertainty Bounding\n(IPUB) method. The IPUB method represents the uncertainties due to missing\nentries as intervals, and efficiently computes the lower and upper bounds of\nthe prediction results when all possible training sets constructed by imputing\narbitrary values in the intervals are considered. The IPUB method can be\napplied to a wide class of convex learning algorithms including penalized\nleast-squares regression, support vector machine (SVM), and logistic\nregression. We demonstrate the advantages of the IPUB method by comparing it\nwith an existing method in numerical experiment with benchmark datasets.\n",
"title": "Interval-based Prediction Uncertainty Bound Computation in Learning with Missing Values"
}
| null | null | null | null | true | null |
17905
| null |
Default
| null | null |
null |
{
"abstract": " This paper provides an entry point to the problem of interpreting a deep\nneural network model and explaining its predictions. It is based on a tutorial\ngiven at ICASSP 2017. It introduces some recently proposed techniques of\ninterpretation, along with theory, tricks and recommendations, to make most\nefficient use of these techniques on real data. It also discusses a number of\npractical applications.\n",
"title": "Methods for Interpreting and Understanding Deep Neural Networks"
}
| null | null | null | null | true | null |
17906
| null |
Default
| null | null |
null |
{
"abstract": " In this digital era, one thing that still holds the convention is a printed\narchive. Printed documents find their use in many critical domains such as\ncontract papers, legal tenders and proof of identity documents. As more\nadvanced printing, scanning and image editing techniques are becoming\navailable, forgeries on these legal tenders pose a serious threat. Ability to\neasily and reliably identify source printer of a printed document can help a\nlot in reducing this menace. During printing procedure, printer hardware\nintroduces certain distortions in printed characters' locations and shapes\nwhich are invisible to naked eyes. These distortions are referred as geometric\ndistortions, their profile (or signature) is generally unique for each printer\nand can be used for printer classification purpose. This paper proposes a set\nof features for characterizing text-line-level geometric distortions, referred\nas geometric distortion signatures and presents a novel system to use them for\nidentification of the origin of a printed document. Detailed experiments\nperformed on a set of thirteen printers demonstrate that the proposed system\nachieves state of the art performance and gives much higher accuracy under\nsmall training size constraint. For four training and six test pages of three\ndifferent fonts, the proposed method gives 99\\% classification accuracy.\n",
"title": "Passive Classification of Source Printer using Text-line-level Geometric Distortion Signatures from Scanned Images of Printed Documents"
}
| null | null | null | null | true | null |
17907
| null |
Default
| null | null |
null |
{
"abstract": " This paper proposes a gamma process for modelling the damage that accumulates\nover time in the lumber used in structural engineering applications when stress\nis applied. The model separates the stochastic processes representing features\ninternal to the piece of lumber on the one hand, from those representing\nexternal forces due to applied dead and live loads. The model applies those\nexternal forces through a time-varying population level function designed for\ntime-varying loads. The application of this type of model, which is standard in\nreliability analysis, is novel in this context, which has been dominated by\naccumulated damage models (ADMs) over more than half a century. The proposed\nmodel is compared with one of the traditional ADMs. Our statistical results\nbased on a Bayesian analysis of experimental data highlight the limitations of\nusing accelerated testing data to assess long-term reliability, as seen in the\nwide posterior intervals. This suggests the need for more comprehensive testing\nin future applications, or to encode appropriate expert knowledge in the priors\nused for Bayesian analysis.\n",
"title": "The duration of load effect in lumber as stochastic degradation"
}
| null | null | null | null | true | null |
17908
| null |
Default
| null | null |
null |
{
"abstract": " Suppose one has data from one or more completed vaccine efficacy trials and\nwishes to estimate the efficacy in a new setting. Often logistical or ethical\nconsiderations make running another efficacy trial impossible. Fortunately, if\nthere is a biomarker that is the primary modifier of efficacy, then the\nbiomarker-conditional efficacy may be identical in the completed trials and the\nnew setting, or at least informative enough to meaningfully bound this\nquantity. Given a sample of this biomarker from the new population, we might\nhope we can bridge the results of the completed trials to estimate the vaccine\nefficacy in this new population. Unfortunately, even knowing the true\nconditional efficacy in the new population fails to identify the marginal\nefficacy due to the unknown conditional unvaccinated risk. We define a curve\nthat partially identifies (lower bounds) the marginal efficacy in the new\npopulation as a function of the population's marginal unvaccinated risk, under\nthe assumption that one can identify bounds on the conditional unvaccinated\nrisk in the new population. Interpreting the curve only requires identifying\nplausible regions of the marginal unvaccinated risk in the new population. We\npresent a nonparametric estimator of this curve and develop valid lower\nconfidence bounds that concentrate at a parametric rate. We use vaccine\nterminology throughout, but the results apply to general binary interventions\nand bounded outcomes.\n",
"title": "Partial Bridging of Vaccine Efficacy to New Populations"
}
| null | null | null | null | true | null |
17909
| null |
Default
| null | null |
null |
{
"abstract": " Let $B{ aut}_1X$ be the Dold-Lashof classifying space of orientable\nfibrations with fiber $X$. For a rationally weakly trivial map $f:X\\to Y$, our\nstrictly induced map $a_f: (Baut_1X)_0\\to (Baut_1Y)_0$ induces a natural map\nfrom a $X_0$-fibration to a $Y_0$-fibration. It is given by a map between the\ndifferential graded Lie algebras of derivations of Sullivan models. We note\nsome conditions that the map $a_f$ admits a section and note some relations\nwith the Halperin conjecture. Furthermore we give the obstruction class for a\nlifting of a classifying map $h: B\\to (Baut_1Y)_0$ and apply it for liftings of\n$G$-actions on $Y$ for a compact connected Lie group $G$ as the case of $B=BG$\nand evaluating of rational toral ranks as $r_0(Y)\\leq r_0(X)$.\n",
"title": "An induced map between rationalized classifying spaces for fibrations"
}
| null | null | null | null | true | null |
17910
| null |
Default
| null | null |
null |
{
"abstract": " Sophisticated gated recurrent neural network architectures like LSTMs and\nGRUs have been shown to be highly effective in a myriad of applications. We\ndevelop an un-gated unit, the statistical recurrent unit (SRU), that is able to\nlearn long term dependencies in data by only keeping moving averages of\nstatistics. The SRU's architecture is simple, un-gated, and contains a\ncomparable number of parameters to LSTMs; yet, SRUs perform favorably to more\nsophisticated LSTM and GRU alternatives, often outperforming one or both in\nvarious tasks. We show the efficacy of SRUs as compared to LSTMs and GRUs in an\nunbiased manner by optimizing respective architectures' hyperparameters in a\nBayesian optimization scheme for both synthetic and real-world tasks.\n",
"title": "The Statistical Recurrent Unit"
}
| null | null | null | null | true | null |
17911
| null |
Default
| null | null |
null |
{
"abstract": " We study many-body localization properties of the disordered XXZ spin chain\nin the Ising phase. Disorder is introduced via a random magnetic field in the\n$z$-direction. We prove a strong form of dynamical exponential clustering for\neigenstates in the droplet spectrum: For any pair of local observables\nseparated by a distance $\\ell$, the sum of the associated correlators over\nthese states decays exponentially in $\\ell$, in expectation. This exponential\nclustering persists under the time evolution in the droplet spectrum. Our\nresult applies to the large disorder regime as well as to the strong Ising\nphase at fixed disorder, with bounds independent of the support of the\nobservables.\n",
"title": "Many-body localization in the droplet spectrum of the random XXZ quantum spin chain"
}
| null | null | null | null | true | null |
17912
| null |
Default
| null | null |
null |
{
"abstract": " Deep neural networks show great potential as solutions to many sensing\napplication problems, but their excessive resource demand slows down execution\ntime, pausing a serious impediment to deployment on low-end devices. To address\nthis challenge, recent literature focused on compressing neural network size to\nimprove performance. We show that changing neural network size does not\nproportionally affect performance attributes of interest, such as execution\ntime. Rather, extreme run-time nonlinearities exist over the network\nconfiguration space. Hence, we propose a novel framework, called FastDeepIoT,\nthat uncovers the non-linear relation between neural network structure and\nexecution time, then exploits that understanding to find network configurations\nthat significantly improve the trade-off between execution time and accuracy on\nmobile and embedded devices. FastDeepIoT makes two key contributions. First,\nFastDeepIoT automatically learns an accurate and highly interpretable execution\ntime model for deep neural networks on the target device. This is done without\nprior knowledge of either the hardware specifications or the detailed\nimplementation of the used deep learning library. Second, FastDeepIoT informs a\ncompression algorithm how to minimize execution time on the profiled device\nwithout impacting accuracy. We evaluate FastDeepIoT using three different\nsensing-related tasks on two mobile devices: Nexus 5 and Galaxy Nexus.\nFastDeepIoT further reduces the neural network execution time by $48\\%$ to\n$78\\%$ and energy consumption by $37\\%$ to $69\\%$ compared with the\nstate-of-the-art compression algorithms.\n",
"title": "FastDeepIoT: Towards Understanding and Optimizing Neural Network Execution Time on Mobile and Embedded Devices"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17913
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, two robust model predictive control (MPC) schemes are proposed\nfor tracking control of nonholonomic systems with bounded disturbances:\ntube-MPC and nominal robust MPC (NRMPC). In tube-MPC, the control signal\nconsists of a control action and a nonlinear feedback law based on the\ndeviation of the actual states from the states of a nominal system. It renders\nthe actual trajectory within a tube centered along the optimal trajectory of\nthe nominal system. Recursive feasibility and input-to-state stability are\nestablished and the constraints are ensured by tightening the input domain and\nthe terminal region. While in NRMPC, an optimal control sequence is obtained by\nsolving an optimization problem based on the current state, and the first\nportion of this sequence is applied to the real system in an open-loop manner\nduring each sampling period. The state of nominal system model is updated by\nthe actual state at each step, which provides additional a feedback. By\nintroducing a robust state constraint and tightening the terminal region,\nrecursive feasibility and input-to-state stability are guaranteed. Simulation\nresults demonstrate the effectiveness of both strategies proposed.\n",
"title": "Robust MPC for tracking of nonholonomic robots with additive disturbances"
}
| null | null | null | null | true | null |
17914
| null |
Default
| null | null |
null |
{
"abstract": " Hidden Markov Models (HMMs) are a ubiquitous tool to model time series data,\nand have been widely used in two main tasks of Automatic Music Transcription\n(AMT): note segmentation, i.e. identifying the played notes after a multi-pitch\nestimation, and sequential post-processing, i.e. correcting note segmentation\nusing training data. In this paper, we employ the multi-pitch estimation method\ncalled Probabilistic Latent Component Analysis (PLCA), and develop AMT systems\nby integrating different HMM-based modules in this framework. For note\nsegmentation, we use two different twostate on/o? HMMs, including a\nhigher-order one for duration modeling. For sequential post-processing, we\nfocused on a musicological modeling of polyphonic harmonic transitions, using a\nfirst- and second-order HMMs whose states are defined through candidate note\nmixtures. These different PLCA plus HMM systems have been evaluated\ncomparatively on two different instrument repertoires, namely the piano (using\nthe MAPS database) and the marovany zither. Our results show that the use of\nHMMs could bring noticeable improvements to transcription results, depending on\nthe instrument repertoire.\n",
"title": "Investigation on the use of Hidden-Markov Models in automatic transcription of music"
}
| null | null | null | null | true | null |
17915
| null |
Default
| null | null |
null |
{
"abstract": " Network growth processes can be understood as generative models of the\nstructure and history of complex networks. This point of view naturally leads\nto the problem of network archaeology: Reconstructing all the past states of a\nnetwork from its structure---a difficult permutation inference problem. In this\npaper, we introduce a Bayesian formulation of network archaeology, with a\ngeneralization of preferential attachment as our generative mechanism. We\ndevelop a sequential importance sampling algorithm to evaluate the posterior\naverages of this model, as well as an efficient heuristic that uncovers the\nhistory of a network in linear time. We use these methods to identify and\ncharacterize a phase transition in the quality of the reconstructed history,\nwhen they are applied to artificial networks generated by the model itself.\nDespite the existence of a no-recovery phase, we find that non-trivial\ninference is possible in a large portion of the parameter space as well as on\nempirical data.\n",
"title": "Network archaeology: phase transition in the recoverability of network history"
}
| null | null | null | null | true | null |
17916
| null |
Default
| null | null |
null |
{
"abstract": " Advances in data analytics bring with them civil rights implications.\nData-driven and algorithmic decision making increasingly determine how\nbusinesses target advertisements to consumers, how police departments monitor\nindividuals or groups, how banks decide who gets a loan and who does not, how\nemployers hire, how colleges and universities make admissions and financial aid\ndecisions, and much more. As data-driven decisions increasingly affect every\ncorner of our lives, there is an urgent need to ensure they do not become\ninstruments of discrimination, barriers to equality, threats to social justice,\nand sources of unfairness. In this paper, we argue for a concrete research\nagenda aimed at addressing these concerns, comprising five areas of emphasis:\n(i) Determining if models and modeling procedures exhibit objectionable bias;\n(ii) Building awareness of fairness into machine learning methods; (iii)\nImproving the transparency and control of data- and model-driven decision\nmaking; (iv) Looking beyond the algorithm(s) for sources of bias and\nunfairness-in the myriad human decisions made during the problem formulation\nand modeling process; and (v) Supporting the cross-disciplinary scholarship\nnecessary to do all of that well.\n",
"title": "Big Data, Data Science, and Civil Rights"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17917
| null |
Validated
| null | null |
null |
{
"abstract": " The present paper is dedicated to the global well-posedness issue for the\nBoussinesq system with the temperature-dependent viscosity in $\\mathbb{R}^2.$\nWe aim at extending the work by Abidi and Zhang ( Adv. Math. 2017 (305)\n1202--1249 ) to a supercritical dissipation for temperature.\n",
"title": "Global well-posedness for 2-D Boussinesq system with the temperature-dependent viscosity and supercritical dissipation"
}
| null | null | null | null | true | null |
17918
| null |
Default
| null | null |
null |
{
"abstract": " We present large-field (4.25~$\\times$~3.75 deg$^2$) mapping observations\ntoward the Galactic region centered at $l = 150\\arcdeg, b = 3.5\\arcdeg$ in the\n$J = 1-0$ emission line of CO isotopologues ($^{12}$CO, $^{13}$CO, and\nC$^{18}$O), using the 13.7 m millimeter-wavelength telescope of the Purple\nMountain Observatory. Based on the $^{13}$CO observations, we reveal a\nfilamentary cloud in the Local Arm at a velocity range of $-$0.5 to\n6.5~km~s$^{-1}$. This molecular cloud contains 1 main filament and 11\nsub-filaments, showing the so-called \"ridge-nest\" structure. The main filament\nand three sub-filaments are also detected in the C$^{18}$O line. The velocity\nstructures of most identified filaments display continuous distribution with\nslight velocity gradients. The measured median excitation temperature, line\nwidth, length, width, and linear mass of the filaments are $\\sim$9.28~K,\n0.85~km~s$^{-1}$, 7.30~pc, 0.79~pc, and 17.92~$M_\\sun$~pc$^{-1}$, respectively,\nassuming a distance of 400~pc. We find that the four filaments detected in the\nC$^{18}$O line are thermally supercritical, and two of them are in the\nvirialized state, and thus tend to be gravitationally bound. We identify in\ntotal 146 $^{13}$CO clumps in the cloud, about 77$\\%$ of the clumps are\ndistributed along the filaments. About 56$\\%$ of the virialized clumps are\nfound to be associated with the supercritical filaments. Three young stellar\nobject (YSO) candidates are also identified in the supercritical filaments,\nbased on the complementary infrared (IR) data. These results indicate that the\nsupercritical filaments, especially the virialized filaments, may contain\nstar-forming activities.\n",
"title": "CO~($J = 1-0$) Observations of a Filamentary Molecular Cloud in the Galactic Region Centered at $l = 150\\arcdeg, b = 3.5\\arcdeg$"
}
| null | null | null | null | true | null |
17919
| null |
Default
| null | null |
null |
{
"abstract": " Unseen data conditions can inflict serious performance degradation on systems\nrelying on supervised machine learning algorithms. Because data can often be\nunseen, and because traditional machine learning algorithms are trained in a\nsupervised manner, unsupervised adaptation techniques must be used to adapt the\nmodel to the unseen data conditions. However, unsupervised adaptation is often\nchallenging, as one must generate some hypothesis given a model and then use\nthat hypothesis to bootstrap the model to the unseen data conditions.\nUnfortunately, reliability of such hypotheses is often poor, given the mismatch\nbetween the training and testing datasets. In such cases, a model hypothesis\nconfidence measure enables performing data selection for the model adaptation.\nUnderlying this approach is the fact that for unseen data conditions, data\nvariability is introduced to the model, which the model propagates to its\noutput decision, impacting decision reliability. In a fully connected network,\nthis data variability is propagated as distortions from one layer to the next.\nThis work aims to estimate the propagation of such distortion in the form of\nnetwork activation entropy, which is measured over a short- time running window\non the activation from each neuron of a given hidden layer, and these\nmeasurements are then used to compute summary entropy. This work demonstrates\nthat such an entropy measure can help to select data for unsupervised model\nadaptation, resulting in performance gains in speech recognition tasks. Results\nfrom standard benchmark speech recognition tasks show that the proposed\napproach can alleviate the performance degradation experienced under unseen\ndata conditions by iteratively adapting the model to the unseen datas acoustic\ncondition.\n",
"title": "Leveraging Deep Neural Network Activation Entropy to cope with Unseen Data in Speech Recognition"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17920
| null |
Validated
| null | null |
null |
{
"abstract": " We study the generation of the matter-antimatter asymmetry during bosonic\npreheating, focusing on the sources of the asymmetry. If the asymmetry appears\nin the multiplication factor of the resonant particle production, the\nmatter-antimatter ratio will grow during preheating. On the other hand, if the\nasymmetry does not grow during preheating, one has to find out another reason.\nWe consider several scenarios for the asymmetric preheating to distinguish the\nsources of the asymmetry. We also discuss a new baryogenesis scenario, in which\nthe asymmetry is generated without introducing neither loop corrections nor\nrotation of a field.\n",
"title": "Asymmetric Preheating"
}
| null | null | null | null | true | null |
17921
| null |
Default
| null | null |
null |
{
"abstract": " This paper introduces the acoustic scene classification task of DCASE 2018\nChallenge and the TUT Urban Acoustic Scenes 2018 dataset provided for the task,\nand evaluates the performance of a baseline system in the task. As in previous\nyears of the challenge, the task is defined for classification of short audio\nsamples into one of predefined acoustic scene classes, using a supervised,\nclosed-set classification setup. The newly recorded TUT Urban Acoustic Scenes\n2018 dataset consists of ten different acoustic scenes and was recorded in six\nlarge European cities, therefore it has a higher acoustic variability than the\nprevious datasets used for this task, and in addition to high-quality binaural\nrecordings, it also includes data recorded with mobile devices. We also present\nthe baseline system consisting of a convolutional neural network and its\nperformance in the subtasks using the recommended cross-validation setup.\n",
"title": "A multi-device dataset for urban acoustic scene classification"
}
| null | null | null | null | true | null |
17922
| null |
Default
| null | null |
null |
{
"abstract": " We propose a class of Particle-In-Cell (PIC) methods for the Vlasov-Poisson\nsystem with a strong and inhomogeneous external magnetic field with fixed\ndirection, where we focus on the motion of particles in the plane orthogonal to\nthe magnetic field (so-called poloidal directions). In this regime, the time\nstep can be subject to stability constraints related to the smallness of Larmor\nradius and plasma frequency. To avoid this limitation, our approach is based on\nfirst and higher-order semi-implicit numerical schemes already validated on\ndissipative systems [3] and for homogeneous magnetic fields [10]. Thus, when\nthe magnitude of the external magnetic field becomes large, this method\nprovides a consistent PIC discretization of the guiding-center system taking\ninto account variations of the magnetic field. We carry out some theoretical\nproofs and perform several numerical experiments that establish a solid\nvalidation of the method and its underlying concepts.\n",
"title": "Asymptotically preserving particle-in-cell methods for inhomogenous strongly magnetized plasmas"
}
| null | null | null | null | true | null |
17923
| null |
Default
| null | null |
null |
{
"abstract": " Methodological research rarely generates a broad interest, yet our work on\nthe validity of cluster inference methods for functional magnetic resonance\nimaging (fMRI) created intense discussion on both the minutia of our approach\nand its implications for the discipline. In the present work, we take on\nvarious critiques of our work and further explore the limitations of our\noriginal work. We address issues about the particular event-related designs we\nused, considering multiple event types and randomisation of events between\nsubjects. We consider the lack of validity found with one-sample permutation\n(sign flipping) tests, investigating a number of approaches to improve the\nfalse positive control of this widely used procedure. We found that the\ncombination of a two-sided test and cleaning the data using ICA FIX resulted in\nnominal false positive rates for all datasets, meaning that data cleaning is\nnot only important for resting state fMRI, but also for task fMRI. Finally, we\ndiscuss the implications of our work on the fMRI literature as a whole,\nestimating that at least 10% of the fMRI studies have used the most problematic\ncluster inference method (P = 0.01 cluster defining threshold), and how\nindividual studies can be interpreted in light of our findings. These\nadditional results underscore our original conclusions, on the importance of\ndata sharing and thorough evaluation of statistical methods on realistic null\ndata.\n",
"title": "Cluster Failure Revisited: Impact of First Level Design and Data Quality on Cluster False Positive Rates"
}
| null | null | null | null | true | null |
17924
| null |
Default
| null | null |
null |
{
"abstract": " Deep learning is an established framework for learning hierarchical data\nrepresentations. While compute power is in abundance, one of the main\nchallenges in applying this framework to robotic grasping has been obtaining\nthe amount of data needed to learn these representations, and structuring the\ndata to the task at hand. Among contemporary approaches in the literature, we\nhighlight key properties that have encouraged the use of deep learning\ntechniques, and in this paper, detail our experience in developing a simulator\nfor collecting cylindrical precision grasps of a multi-fingered dexterous\nrobotic hand.\n",
"title": "An Integrated Simulator and Dataset that Combines Grasping and Vision for Deep Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17925
| null |
Validated
| null | null |
null |
{
"abstract": " Demographic studies suggest that changes in the retinal vasculature geometry,\nespecially in vessel width, are associated with the incidence or progression of\neye-related or systemic diseases. To date, the main information source for\nwidth estimation from fundus images has been the intensity profile between\nvessel edges. However, there are many factors affecting the intensity profile:\npathologies, the central light reflex and local illumination levels, to name a\nfew. In this study, we introduce three information sources for width\nestimation. These are the probability profiles of vessel interior, centreline\nand edge locations generated by a deep network. The probability profiles\nprovide direct access to vessel geometry and are used in the likelihood\ncalculation for a Bayesian method, particle filtering. We also introduce a\ngeometric model which can handle non-ideal conditions of the probability\nprofiles. Our experiments conducted on the REVIEW dataset yielded consistent\nestimates of vessel width, even in cases when one of the vessel edges is\ndifficult to identify. Moreover, our results suggest that the method is better\nthan human observers at locating edges of low contrast vessels.\n",
"title": "A Recursive Bayesian Approach To Describe Retinal Vasculature Geometry"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17926
| null |
Validated
| null | null |
null |
{
"abstract": " In this Review we will study rigorously the notion of mixed states and their\ndensity matrices. We mostly give complete proofs. We will also discuss the\nquantum-mechanical consequences of possible variations of Planck's constant h.\nThis Review has been written having in mind two readerships: mathematical\nphysicists and quantum physicists. The mathematical rigor is maximal, but the\nlanguage and notation we use throughout should be familiar to physicists.\n",
"title": "Quantum Harmonic Analysis of the Density Matrix: Basics"
}
| null | null |
[
"Mathematics"
] | null | true | null |
17927
| null |
Validated
| null | null |
null |
{
"abstract": " Background: The chromatin remodelers of the SWI/SNF family are critical\ntranscriptional regulators. Recognition of lysine acetylation through a\nbromodomain (BRD) component is key to SWI/SNF function; in most eukaryotes,\nthis function is attributed to SNF2/Brg1.\nResults: Using affinity purification coupled to mass spectrometry (AP-MS) we\nidentified members of a SWI/SNF complex (SWI/SNFTt) in Tetrahymena thermophila.\nSWI/SNFTt is composed of 11 proteins, Snf5Tt, Swi1Tt, Swi3Tt, Snf12Tt, Brg1Tt,\ntwo proteins with potential chromatin interacting domains and four proteins\nwithout orthologs to SWI/SNF proteins in yeast or mammals. SWI/SNFTt subunits\nlocalize exclusively to the transcriptionally active macronucleus (MAC) during\ngrowth and development, consistent with a role in transcription. While\nTetrahymena Brg1 does not contain a BRD, our AP-MS results identified a\nBRD-containing SWI/SNFTt component, Ibd1 that associates with SWI/SNFTt during\ngrowth but not development. AP-MS analysis of epitope-tagged Ibd1 revealed it\nto be a subunit of several additional protein complexes, including putative\nSWRTt, and SAGATt complexes as well as a putative H3K4-specific histone methyl\ntransferase complex. Recombinant Ibd1 recognizes acetyl-lysine marks on\nhistones correlated with active transcription. Consistent with our AP-MS and\nhistone array data suggesting a role in regulation of gene expression, ChIP-Seq\nanalysis of Ibd1 indicated that it primarily binds near promoters and within\ngene bodies of highly expressed genes during growth.\nConclusions: Our results suggest that through recognizing specific histones\nmarks, Ibd1 targets active chromatin regions of highly expressed genes in\nTetrahymena where it subsequently might coordinate the recruitment of several\nchromatin remodeling complexes to regulate the transcriptional landscape of\nvegetatively growing Tetrahymena cells.\n",
"title": "The bromodomain-containing protein Ibd1 links multiple chromatin related protein complexes to highly expressed genes in Tetrahymena thermophila"
}
| null | null | null | null | true | null |
17928
| null |
Default
| null | null |
null |
{
"abstract": " We consider two questions at the heart of machine learning; how can we\npredict if a minimum will generalize to the test set, and why does stochastic\ngradient descent find minima that generalize well? Our work responds to Zhang\net al. (2016), who showed deep neural networks can easily memorize randomly\nlabeled training data, despite generalizing well on real labels of the same\ninputs. We show that the same phenomenon occurs in small linear models. These\nobservations are explained by the Bayesian evidence, which penalizes sharp\nminima but is invariant to model parameterization. We also demonstrate that,\nwhen one holds the learning rate fixed, there is an optimum batch size which\nmaximizes the test set accuracy. We propose that the noise introduced by small\nmini-batches drives the parameters towards minima whose evidence is large.\nInterpreting stochastic gradient descent as a stochastic differential equation,\nwe identify the \"noise scale\" $g = \\epsilon (\\frac{N}{B} - 1) \\approx \\epsilon\nN/B$, where $\\epsilon$ is the learning rate, $N$ the training set size and $B$\nthe batch size. Consequently the optimum batch size is proportional to both the\nlearning rate and the size of the training set, $B_{opt} \\propto \\epsilon N$.\nWe verify these predictions empirically.\n",
"title": "A Bayesian Perspective on Generalization and Stochastic Gradient Descent"
}
| null | null | null | null | true | null |
17929
| null |
Default
| null | null |
null |
{
"abstract": " We provide a full analysis of ghost free higher derivative field theories\nwith coupled degrees of freedom. Assuming the absence of gauge symmetries, we\nderive the degeneracy conditions in order to evade the Ostrogradsky ghosts, and\nanalyze which (non)trivial classes of solutions this allows for. It is shown\nexplicitly how Lorentz invariance avoids the propagation of \"half\" degrees of\nfreedom. Moreover, for a large class of theories, we construct the field\nredefinitions and/or (extended) contact transformations that put the theory in\na manifestly first order form. Finally, we identify which class of theories\ncannot be brought to first order form by such transformations.\n",
"title": "Higher Derivative Field Theories: Degeneracy Conditions and Classes"
}
| null | null | null | null | true | null |
17930
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a novel approach for predicting the progression of adolescent\nidiopathic scoliosis from 3D spine models reconstructed from biplanar X-ray\nimages. Recent progress in machine learning have allowed to improve\nclassification and prognosis rates, but lack a probabilistic framework to\nmeasure uncertainty in the data. We propose a discriminative probabilistic\nmanifold embedding where locally linear mappings transform data points from\nhigh-dimensional space to corresponding low-dimensional coordinates. A\ndiscriminant adjacency matrix is constructed to maximize the separation between\nprogressive and non-progressive groups of patients diagnosed with scoliosis,\nwhile minimizing the distance in latent variables belonging to the same class.\nTo predict the evolution of deformation, a baseline reconstruction is projected\nonto the manifold, from which a spatiotemporal regression model is built from\nparallel transport curves inferred from neighboring exemplars. Rate of\nprogression is modulated from the spine flexibility and curve magnitude of the\n3D spine deformation. The method was tested on 745 reconstructions from 133\nsubjects using longitudinal 3D reconstructions of the spine, with results\ndemonstrating the discriminatory framework can identify between progressive and\nnon-progressive of scoliotic patients with a classification rate of 81% and\nprediction differences of 2.1$^{o}$ in main curve angulation, outperforming\nother manifold learning methods. Our method achieved a higher prediction\naccuracy and improved the modeling of spatiotemporal morphological changes in\nhighly deformed spines compared to other learning methods.\n",
"title": "3D Morphology Prediction of Progressive Spinal Deformities from Probabilistic Modeling of Discriminant Manifolds"
}
| null | null | null | null | true | null |
17931
| null |
Default
| null | null |
null |
{
"abstract": " Many of the baryons in our Galaxy probably lie outside the well known disk\nand bulge components. Despite a wealth of evidence for the presence of some gas\nin galactic halos, including absorption line systems in the spectra of quasars,\nhigh velocity neutral hydrogen clouds in our Galaxy halo, line emitting ionised\nhydrogen originating from galactic winds in nearby starburst galaxies, and the\nX-ray coronas surrounding the most massive galaxies, accounting for the gas in\nthe halo of any galaxy has been observationally challenging primarily because\nof its low density in the expansive halo. The most sensitive measurements come\nfrom detecting absorption by the intervening gas in the spectra of distant\nobjects such as quasars or distant halo stars, but these have typically been\nlimited to a few lines of sight to sufficiently bright objects. Massive\nspectroscopic surveys of millions of objects provide an alternative approach to\nthe problem. Here, we present the first evidence for a widely distributed,\nneutral, excited hydrogen component of the Galaxy's halo. It is observed as the\nslight, (0.779 $\\pm$ 0.006)\\%, absorption of flux near the rest wavelength of\nH$\\alpha$ in the combined spectra of hundreds of thousands of galaxy spectra\nand is ubiquitous in high latitude lines of sight. This observation provides an\navenue to tracing, both spatially and kinematically, the majority of the gas in\nthe halo of our Galaxy.\n",
"title": "The Galaxy's Veil of Excited Hydrogen"
}
| null | null |
[
"Physics"
] | null | true | null |
17932
| null |
Validated
| null | null |
null |
{
"abstract": " Supercontinuum generation using chip-integrated photonic waveguides is a\npowerful approach for spectrally broadening pulsed laser sources with very low\npulse energies and compact form factors. When pumped with a mode-locked laser\nfrequency comb, these waveguides can coherently expand the comb spectrum to\nmore than an octave in bandwidth to enable self-referenced stabilization.\nHowever, for applications in frequency metrology and precision spectroscopy, it\nis desirable to not only support self-referencing, but also to generate\nlow-noise combs with customizable broadband spectra. In this work, we\ndemonstrate dispersion-engineered waveguides based on silicon nitride that are\ndesigned to meet these goals and enable precision optical metrology experiments\nacross large wavelength spans. We perform a clock comparison measurement and\nreport a clock-limited relative frequency instability of $3.8\\times10^{-15}$ at\n$\\tau = 2$ seconds between a 1550 nm cavity-stabilized reference laser and\nNIST's calcium atomic clock laser at 657 nm using a two-octave\nwaveguide-supercontinuum comb.\n",
"title": "Photonic-chip supercontinuum with tailored spectra for precision frequency metrology"
}
| null | null | null | null | true | null |
17933
| null |
Default
| null | null |
null |
{
"abstract": " Through a series of examples, we illustrate some important drawbacks that the\naction logic framework suffers from in its ability to represent the dynamics of\ninformation updates. We argue that these problems stem from the fact that the\naction model, a central construct designed to encode agents' uncertainty about\nactions, is itself effectively common knowledge amongst the agents. In response\nto these difficulties, we motivate and propose an alternative semantics that\navoids them by (roughly speaking) endogenizing the action model. We discuss the\nrelationship to action logic, and provide a sound and complete axiomatization.\n",
"title": "Endogenizing Epistemic Actions"
}
| null | null | null | null | true | null |
17934
| null |
Default
| null | null |
null |
{
"abstract": " We give a rigorous characterization of what it means for a programming\nlanguage to be memory safe, capturing the intuition that memory safety supports\nlocal reasoning about state. We formalize this principle in two ways. First, we\nshow how a small memory-safe language validates a noninterference property: a\nprogram can neither affect nor be affected by unreachable parts of the state.\nSecond, we extend separation logic, a proof system for heap-manipulating\nprograms, with a memory-safe variant of its frame rule. The new rule is\nstronger because it applies even when parts of the program are buggy or\nmalicious, but also weaker because it demands a stricter form of separation\nbetween parts of the program state. We also consider a number of pragmatically\nmotivated variations on memory safety and the reasoning principles they\nsupport. As an application of our characterization, we evaluate the security of\na previously proposed dynamic monitor for memory safety of heap-allocated data.\n",
"title": "The Meaning of Memory Safety"
}
| null | null | null | null | true | null |
17935
| null |
Default
| null | null |
null |
{
"abstract": " We present a new class of service for location based social networks, called\nthe Flexible Group Spatial Keyword Query, which enables a group of users to\ncollectively find a point of interest (POI) that optimizes an aggregate cost\nfunction combining both spatial distances and keyword similarities. In\naddition, our query service allows users to consider the tradeoffs between\nobtaining a sub-optimal solution for the entire group and obtaining an\noptimimized solution but only for a subgroup.\nWe propose algorithms to process three variants of the query: (i) the group\nnearest neighbor with keywords query, which finds a POI that optimizes the\naggregate cost function for the whole group of size n, (ii) the subgroup\nnearest neighbor with keywords query, which finds the optimal subgroup and a\nPOI that optimizes the aggregate cost function for a given subgroup size m (m\n<= n), and (iii) the multiple subgroup nearest neighbor with keywords query,\nwhich finds optimal subgroups and corresponding POIs for each of the subgroup\nsizes in the range [m, n]. We design query processing algorithms based on\nbranch-and-bound and best-first paradigms. Finally, we provide theoretical\nbounds and conduct extensive experiments with two real datasets which verify\nthe effectiveness and efficiency of the proposed algorithms.\n",
"title": "The Flexible Group Spatial Keyword Query"
}
| null | null | null | null | true | null |
17936
| null |
Default
| null | null |
null |
{
"abstract": " We characterize the completeness and frame/basis property of a union of\nunder-sampled windowed exponentials of the form $$ {\\mathcal F}(g): =\\{e^{2\\pi\ni n x}: n\\ge 0\\}\\cup \\{g(x)e^{2\\pi i nx}: n<0\\} $$ for $L^2[-1/2,1/2]$ by the\nspectra of the Toeplitz operators with symbol $g$. Using this characterization,\nwe classify all real-valued functions $g$ such that ${\\mathcal F}(g)$ is\ncomplete or forms a frame/basis. Conversely, we use the classical\nKadec-1/4-theorem in non-harmonic Fourier series to determine all $\\xi$ such\nthat the Toeplitz operators with symbol $e^{2\\pi i \\xi x}$ is injective or\ninvertible. These results demonstrate an elegant interaction between frame\ntheory of windowed exponentials and Toeplitz operators. Finally, as an\napplication, we use our results to answer some open questions in dynamical\nsampling, phase retrieval and derivative samplings on $\\ell^2({\\mathbb Z})$ and\nPaley-Wiener spaces of bandlimited functions.\n",
"title": "Undersampled windowed exponentials and their applications"
}
| null | null | null | null | true | null |
17937
| null |
Default
| null | null |
null |
{
"abstract": " Deep learning models are often successfully trained using gradient descent,\ndespite the worst case hardness of the underlying non-convex optimization\nproblem. The key question is then under what conditions can one prove that\noptimization will succeed. Here we provide a strong result of this kind. We\nconsider a neural net with one hidden layer and a convolutional structure with\nno overlap and a ReLU activation function. For this architecture we show that\nlearning is NP-complete in the general case, but that when the input\ndistribution is Gaussian, gradient descent converges to the global optimum in\npolynomial time. To the best of our knowledge, this is the first global\noptimality guarantee of gradient descent on a convolutional neural network with\nReLU activations.\n",
"title": "Globally Optimal Gradient Descent for a ConvNet with Gaussian Inputs"
}
| null | null | null | null | true | null |
17938
| null |
Default
| null | null |
null |
{
"abstract": " Computer based recognition and detection of abnormalities in ECG signals is\nproposed. For this purpose, the Support Vector Machines (SVM) are combined with\nthe advantages of Hermite transform representation. SVM represent a special\ntype of classification techniques commonly used in medical applications.\nAutomatic classification of ECG could make the work of cardiologic departments\nfaster and more efficient. It would also reduce the number of false diagnosis\nand, as a result, save lives. The working principle of the SVM is based on\ntranslating the data into a high dimensional feature space and separating it\nusing a linear classificator. In order to provide an optimal representation for\nSVM application, the Hermite transform domain is used. This domain is proved to\nbe suitable because of the similarity of the QRS complex with Hermite basis\nfunctions. The maximal signal information is obtained using a small set of\nfeatures that are used for detection of irregular QRS complexes. The aim of the\npaper is to show that these features can be employed for automatic ECG signal\nanalysis.\n",
"title": "Detection of irregular QRS complexes using Hermite Transform and Support Vector Machine"
}
| null | null | null | null | true | null |
17939
| null |
Default
| null | null |
null |
{
"abstract": " Let $A$ be an expanding $d\\times d$ matrix with integer entries and\n${\\mathcal D}\\subset {\\mathbb Z}^d$ be a finite digit set. Then the pair $(A,\n{\\mathcal D})$ defines a unique integral self-affine set $K=A^{-1}(K+{\\mathcal\nD})$. In this paper, by replacing the Euclidean norm with a pseudo-norm $w$ in\nterms of $A$, we construct a hyperbolic graph on $(A, {\\mathcal D})$ and show\nthat $K$ can be identified with the hyperbolic boundary. Moreover, if $(A,\n{\\mathcal D})$ safisfies the open set condition, we also prove that two totally\ndisconnected integral self-affine sets are Lipschitz equivalent if an only if\nthey have the same $w$-Hausdorff dimension, that is, their digit sets have\nequal cardinality. We extends some well-known results in the self-similar sets\nto the self-affine sets.\n",
"title": "On the Lipschitz equivalence of self-affine sets"
}
| null | null | null | null | true | null |
17940
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we consider the Dvali and Gómez assumption that the end state\nof a gravitational collapse is a Bose-Einstein condensate of gravitons. We then\nconstruct the two Gross-Pitaevskii equations for a static and spherically\nsymmetric configuration of the condensate. These two equations correspond to\nthe constrained minimisation of the gravitational Hamiltonian with respect to\nthe redshift and the Newtonian potential, per given number of gravitons. We\nfind that the effective geometry of the condensate is the one of a gravastar (a\nDeSitter star) with a sub-Planckian cosmological constant, for masses larger\nthan the Planck scale. Thus, a condensate corresponding to a semiclassical\nblack hole, is always quantum and weakly coupled. Finally, we obtain that the\nboundary of our gravastar, although it is not the location of a horizon,\ncorresponds to the Schwarzschild radius.\n",
"title": "The Gross-Pitaevskii equations of a static and spherically symmetric condensate of gravitons"
}
| null | null | null | null | true | null |
17941
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of convergence to a saddle point of a concave-convex\nfunction via gradient dynamics. Since first introduced by Arrow, Hurwicz and\nUzawa in [1] such dynamics have been extensively used in diverse areas, there\nare, however, features that render their analysis non trivial. These include\nthe lack of convergence guarantees when the function considered is not strictly\nconcave-convex and also the non-smoothness of subgradient dynamics. Our aim in\nthis two part paper is to provide an explicit characterization to the\nasymptotic behaviour of general gradient and subgradient dynamics applied to a\ngeneral concave-convex function. We show that despite the nonlinearity and\nnon-smoothness of these dynamics their $\\omega$-limit set is comprised of\ntrajectories that solve only explicit linear ODEs that are characterized within\nthe paper.\nMore precisely, in Part I an exact characterization is provided to the\nasymptotic behaviour of unconstrained gradient dynamics. We also show that when\nconvergence to a saddle point is not guaranteed then the system behaviour can\nbe problematic, with arbitrarily small noise leading to an unbounded variance.\nIn Part II we consider a general class of subgradient dynamics that restrict\ntrajectories in an arbitrary convex domain, and show that their limiting\ntrajectories are solutions of subgradient dynamics on only affine subspaces.\nThe latter is a smooth class of dynamics with an asymptotic behaviour exactly\ncharacterized in Part I, as solutions to explicit linear ODEs. These results\nare used to formulate corresponding convergence criteria and are demonstrated\nwith several examples and applications presented in Part II.\n",
"title": "Stability and instability in saddle point dynamics - Part I"
}
| null | null | null | null | true | null |
17942
| null |
Default
| null | null |
null |
{
"abstract": " We investigate to what extent mobile use patterns can predict -- at the\nmoment it is posted -- whether a notification will be clicked within the next\n10 minutes. We use a data set containing the detailed mobile phone usage logs\nof 279 users, who over the course of 5 weeks received 446,268 notifications\nfrom a variety of apps. Besides using classical gradient-boosted trees, we\ndemonstrate how to make continual predictions using a recurrent neural network\n(RNN). The two approaches achieve a similar AUC of ca. 0.7 on unseen users,\nwith a possible operation point of 50% sensitivity and 80% specificity\nconsidering all notification types (an increase of 40% with respect to a\nprobabilistic baseline). These results enable automatic, intelligent handling\nof mobile phone notifications without the need for user feedback or\npersonalization. Furthermore, they showcase how forego feature-extraction by\nusing RNNs for continual predictions directly on mobile usage logs. To the best\nof our knowledge, this is the first work that leverages mobile sensor data for\ncontinual, context-aware predictions of interruptibility using deep neural\nnetworks.\n",
"title": "Continual Prediction of Notification Attendance with Classical and Deep Network Approaches"
}
| null | null | null | null | true | null |
17943
| null |
Default
| null | null |
null |
{
"abstract": " In a previous study, the algebraic formulation of the First Fundamental\nTheorem of Calculus (FFTC) is shown to allow extensions of differential and\nRota-Baxter operators on the one hand, and to give rise to liftings of monads\nand comonads, and mixed distributive laws on the other. Generalizing the FFTC,\nwe consider in this paper a class of constraints between a differential\noperator and a Rota-Baxter operator. For a given constraint, we show that the\nexistences of extensions of differential and Rota-Baxter operators, of liftings\nof monads and comonads, and of mixed distributive laws are equivalent. We\nfurther give a classification of the constraints satisfying these equivalent\nconditions.\n",
"title": "Extensions of Operators, Liftings of Monads and Distributive Laws"
}
| null | null | null | null | true | null |
17944
| null |
Default
| null | null |
null |
{
"abstract": " Micro Aerial Vehicles (MAVs) are limited in their operation outdoors near\nobstacles by their ability to withstand wind gusts. Currently widespread\nposition control methods such as Proportional Integral Derivative control do\nnot perform well under the influence of gusts. Incremental Nonlinear Dynamic\nInversion (INDI) is a sensor-based control technique that can control nonlinear\nsystems subject to disturbances. It was developed for the attitude control of\nmanned aircraft or MAVs. In this paper we generalize this method to the outer\nloop control of MAVs under severe gust loads. Significant improvements over a\ntraditional Proportional Integral Derivative (PID) controller are demonstrated\nin an experiment where the quadrotor flies in and out of a windtunnel exhaust\nat 10 m/s. The control method does not rely on frequent position updates, as is\ndemonstrated in an outside experiment using a standard GPS module. Finally, we\ninvestigate the effect of using a linearization to calculate thrust vector\nincrements, compared to a nonlinear calculation. The method requires little\nmodeling and is computationally efficient.\n",
"title": "Cascaded Incremental Nonlinear Dynamic Inversion Control for MAV Disturbance Rejection"
}
| null | null | null | null | true | null |
17945
| null |
Default
| null | null |
null |
{
"abstract": " We describe algorithms to compute elliptic functions and their relatives\n(Jacobi theta functions, modular forms, elliptic integrals, and the\narithmetic-geometric mean) numerically to arbitrary precision with rigorous\nerror bounds for arbitrary complex variables. Implementations in ball\narithmetic are available in the open source Arb library. We discuss the\nalgorithms from a concrete implementation point of view, with focus on\nperformance at tens to thousands of digits of precision.\n",
"title": "Numerical Evaluation of Elliptic Functions, Elliptic Integrals and Modular Forms"
}
| null | null | null | null | true | null |
17946
| null |
Default
| null | null |
null |
{
"abstract": " Magnetic skyrmions are localized nanometric spin textures with quantized\nwinding numbers as the topological invariant. Rapidly increasing attention has\nbeen paid to the investigations of skyrmions since their experimental discovery\nin 2009, due both to the fundamental properties and the promising potential in\nspintronics based applications. However, controlled creation of skyrmions\nremains a pivotal challenge towards technological applications. Here, we report\nthat skyrmions can be created locally by electric field in the magnetoelectric\nhelimagnet Cu$\\mathsf{_2}$OSeO$\\mathsf{_3}$. Using Lorentz transmission\nelectron microscopy, we successfully write skyrmions in situ from a helical\nspin background. Our discovery is highly coveted since it implies that\nskyrmionics can be integrated into contemporary field effect transistor based\nelectronic technology, where very low energy dissipation can be achieved, and\nhence realizes a large step forward to its practical applications.\n",
"title": "In situ Electric Field Skyrmion Creation in Magnetoelectric Cu$_2$OSeO$_3$"
}
| null | null | null | null | true | null |
17947
| null |
Default
| null | null |
null |
{
"abstract": " Generative models have long been the dominant approach for speech\nrecognition. The success of these models however relies on the use of\nsophisticated recipes and complicated machinery that is not easily accessible\nto non-practitioners. Recent innovations in Deep Learning have given rise to an\nalternative - discriminative models called Sequence-to-Sequence models, that\ncan almost match the accuracy of state of the art generative models. While\nthese models are easy to train as they can be trained end-to-end in a single\nstep, they have a practical limitation that they can only be used for offline\nrecognition. This is because the models require that the entirety of the input\nsequence be available at the beginning of inference, an assumption that is not\nvalid for instantaneous speech recognition. To address this problem, online\nsequence-to-sequence models were recently introduced. These models are able to\nstart producing outputs as data arrives, and the model feels confident enough\nto output partial transcripts. These models, like sequence-to-sequence are\ncausal - the output produced by the model until any time, $t$, affects the\nfeatures that are computed subsequently. This makes the model inherently more\npowerful than generative models that are unable to change features that are\ncomputed from the data. This paper highlights two main contributions - an\nimprovement to online sequence-to-sequence model training, and its application\nto noisy settings with mixed speech from two speakers.\n",
"title": "An online sequence-to-sequence model for noisy speech recognition"
}
| null | null | null | null | true | null |
17948
| null |
Default
| null | null |
null |
{
"abstract": " For hidden Markov models one of the most popular estimates of the hidden\nchain is the Viterbi path -- the path maximising the posterior probability. We\nconsider a more general setting, called the pairwise Markov model, where the\njoint process consisting of finite-state hidden regime and observation process\nis assumed to be a Markov chain. We prove that under some conditions it is\npossible to extend the Viterbi path to infinity for almost every observation\nsequence which in turn enables to define an infinite Viterbi decoding of the\nobservation process, called the Viterbi process. This is done by constructing a\nblock of observations, called a barrier, which ensures that the Viterbi path\ngoes trough a given state whenever this block occurs in the observation\nsequence.\n",
"title": "Existence of infinite Viterbi path for pairwise Markov models"
}
| null | null | null | null | true | null |
17949
| null |
Default
| null | null |
null |
{
"abstract": " This paper considers insertion and deletion channels with the additional\nassumption that the channel input sequence is implicitly divided into segments\nsuch that at most one edit can occur within a segment. No segment markers are\navailable in the received sequence. We propose code constructions for the\nsegmented deletion, segmented insertion, and segmented insertion-deletion\nchannels based on subsets of Varshamov-Tenengolts codes chosen with\npre-determined prefixes and/or suffixes. The proposed codes, constructed for\nany finite alphabet, are zero-error and can be decoded segment-by-segment. We\nalso derive an upper bound on the rate of any zero-error code for the segmented\nedit channel, in terms of the segment length. This upper bound shows that the\nrate scaling of the proposed codes as the segment length increases is the same\nas that of the maximal code.\n",
"title": "Coding for Segmented Edit Channels"
}
| null | null | null | null | true | null |
17950
| null |
Default
| null | null |
null |
{
"abstract": " Periodic supercell models of electric double layers formed at the interface\nbetween a charged surface and an electrolyte are subject to serious finite size\nerrors and require certain adjustments in the treatment of the long-range\nelectrostatic interactions. In a previous publication (C. Zhang, M. Sprik,\nPhys. Rev. B 94, 245309 (2016)) we have shown how this can be achieved using\nfinite field methods. The test system was the familiar simple point charge\nmodel of a NaCl aqueous solution confined between two oppositely charged walls.\nHere this method is extended to the interface between the (111) polar surface\nof a NaCl crystal and a high concentration NaCl aqueous solution. The crystal\nis kept completely rigid and the compensating charge screening the polarization\ncan only be provided by the electrolyte. We verify that the excess electrolyte\nionic charge at the interface conforms to the Tasker 1/2 rule for compensating\ncharge in the theory of polar rocksalt (111) surfaces. The interface can be\nviewed as an electric double layer with a net charge. We define a generalized\nHelmholtz capacitance $C_\\text{H}$ which can be computed by varying the applied\nelectric field. We find $C_\\text{H} = 8.23 \\, \\mu \\mathrm{Fcm}^{-2}$, which\nshould be compared to the $4.23 \\, \\mu \\mathrm{Fcm}^{-2}$ for the (100)\nnon-polar surface of the same NaCl crystal. This is rationalized by the\nobservation that compensating ions shed their first solvation shell adsorbing\nas contact ions pairs on the polar surface.\n",
"title": "Charge compensation at the interface between the polar NaCl(111) surface and a NaCl aqueous solution"
}
| null | null | null | null | true | null |
17951
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we present CrowdTone, a system designed to help people set the\nappropriate tone in their email communication. CrowdTone utilizes the context\nand content of an email message to identify and set the appropriate tone\nthrough a consensus-building process executed by crowd workers. We evaluated\nCrowdTone with 22 participants, who provided a total of 29 emails that they had\nreceived in the past, and ran them through CrowdTone. Participants and\nprofessional writers assessed the quality of improvements finding a substantial\nincrease in the percentage of emails deemed \"appropriate\" or \"very appropriate\"\n- from 25% to more than 90% by recipients, and from 45% to 90% by professional\nwriters. Additionally, the recipients' feedback indicated that more than 90% of\nthe CrowdTone processed emails showed improvement.\n",
"title": "CrowdTone: Crowd-powered tone feedback and improvement system for emails"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17952
| null |
Validated
| null | null |
null |
{
"abstract": " Restricted Boltzmann machines (RBMs) are energy-based neural-networks which\nare commonly used as the building blocks for deep architectures neural\narchitectures. In this work, we derive a deterministic framework for the\ntraining, evaluation, and use of RBMs based upon the Thouless-Anderson-Palmer\n(TAP) mean-field approximation of widely-connected systems with weak\ninteractions coming from spin-glass theory. While the TAP approach has been\nextensively studied for fully-visible binary spin systems, our construction is\ngeneralized to latent-variable models, as well as to arbitrarily distributed\nreal-valued spin systems with bounded support. In our numerical experiments, we\ndemonstrate the effective deterministic training of our proposed models and are\nable to show interesting features of unsupervised learning which could not be\ndirectly observed with sampling. Additionally, we demonstrate how to utilize\nour TAP-based framework for leveraging trained RBMs as joint priors in\ndenoising problems.\n",
"title": "A Deterministic and Generalized Framework for Unsupervised Learning with Restricted Boltzmann Machines"
}
| null | null |
[
"Computer Science",
"Physics",
"Statistics"
] | null | true | null |
17953
| null |
Validated
| null | null |
null |
{
"abstract": " Compound random measures (CoRM's) are a flexible and tractable framework for\nvectors of completely random measure. In this paper, we provide conditions to\nguarantee the existence of a CoRM. Furthermore, we prove some interesting\nproperties of CoRM's when exponential scores and regularly varying Lévy\nintensities are considered.\n",
"title": "Integrability conditions for Compound Random Measures"
}
| null | null | null | null | true | null |
17954
| null |
Default
| null | null |
null |
{
"abstract": " The test of gravitational force on antimatter in the field of the matter\ngravitational field, produced by earth, can be done by a free fall experiment\nwhich involves only General Relativity, and with a Mach-Zehnder interferometer\nwhich involves Quantum Mechanics. This article presents a new method to produce\na tunable low energy (Ps ) beam suitable for trapping the (Hbar + ) ion in a\nfree fall experiment, and suitable for a gravity Mach-Zehnder interferometer\nwith (Ps). The low energy (Ps) beam is tunable in the [10 eV, 100 eV] range.\n",
"title": "Tunable low energy Ps beam for the anti-hydrogen free fall and for testing gravity with a Mach-Zehnder interferometer"
}
| null | null | null | null | true | null |
17955
| null |
Default
| null | null |
null |
{
"abstract": " Attention-based sequence-to-sequence models for automatic speech recognition\njointly train an acoustic model, language model, and alignment mechanism. Thus,\nthe language model component is only trained on transcribed audio-text pairs.\nThis leads to the use of shallow fusion with an external language model at\ninference time. Shallow fusion refers to log-linear interpolation with a\nseparately trained language model at each step of the beam search. In this\nwork, we investigate the behavior of shallow fusion across a range of\nconditions: different types of language models, different decoding units, and\ndifferent tasks. On Google Voice Search, we demonstrate that the use of shallow\nfusion with a neural LM with wordpieces yields a 9.1% relative word error rate\nreduction (WERR) over our competitive attention-based sequence-to-sequence\nmodel, obviating the need for second-pass rescoring.\n",
"title": "An analysis of incorporating an external language model into a sequence-to-sequence model"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17956
| null |
Validated
| null | null |
null |
{
"abstract": " We propose a new approach to train the Generative Adversarial Nets (GANs)\nwith a mixture of generators to overcome the mode collapsing problem. The main\nintuition is to employ multiple generators, instead of using a single one as in\nthe original GAN. The idea is simple, yet proven to be extremely effective at\ncovering diverse data modes, easily overcoming the mode collapse and delivering\nstate-of-the-art results. A minimax formulation is able to establish among a\nclassifier, a discriminator, and a set of generators in a similar spirit with\nGAN. Generators create samples that are intended to come from the same\ndistribution as the training data, whilst the discriminator determines whether\nsamples are true data or generated by generators, and the classifier specifies\nwhich generator a sample comes from. The distinguishing feature is that\ninternal samples are created from multiple generators, and then one of them\nwill be randomly selected as final output similar to the mechanism of a\nprobabilistic mixture model. We term our method Mixture GAN (MGAN). We develop\ntheoretical analysis to prove that, at the equilibrium, the Jensen-Shannon\ndivergence (JSD) between the mixture of generators' distributions and the\nempirical data distribution is minimal, whilst the JSD among generators'\ndistributions is maximal, hence effectively avoiding the mode collapse. By\nutilizing parameter sharing, our proposed model adds minimal computational cost\nto the standard GAN, and thus can also efficiently scale to large-scale\ndatasets. We conduct extensive experiments on synthetic 2D data and natural\nimage databases (CIFAR-10, STL-10 and ImageNet) to demonstrate the superior\nperformance of our MGAN in achieving state-of-the-art Inception scores over\nlatest baselines, generating diverse and appealing recognizable objects at\ndifferent resolutions, and specializing in capturing different types of objects\nby generators.\n",
"title": "Multi-Generator Generative Adversarial Nets"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
17957
| null |
Validated
| null | null |
null |
{
"abstract": " The seminal work of Gatys et al. demonstrated the power of Convolutional\nNeural Networks (CNNs) in creating artistic imagery by separating and\nrecombining image content and style. This process of using CNNs to render a\ncontent image in different styles is referred to as Neural Style Transfer\n(NST). Since then, NST has become a trending topic both in academic literature\nand industrial applications. It is receiving increasing attention and a variety\nof approaches are proposed to either improve or extend the original NST\nalgorithm. In this paper, we aim to provide a comprehensive overview of the\ncurrent progress towards NST. We first propose a taxonomy of current algorithms\nin the field of NST. Then, we present several evaluation methods and compare\ndifferent NST algorithms both qualitatively and quantitatively. The review\nconcludes with a discussion of various applications of NST and open problems\nfor future research. A list of papers discussed in this review, corresponding\ncodes, pre-trained models and more comparison results are publicly available at\nthis https URL.\n",
"title": "Neural Style Transfer: A Review"
}
| null | null | null | null | true | null |
17958
| null |
Default
| null | null |
null |
{
"abstract": " We aim at characterizing the large-scale distribution of H2O in G327.3-0.6, a\nmassive star-forming region made of individual objects in different\nevolutionary phases. We investigate variations of H2O abundance as function of\nevolution. We present Herschel continuum maps at 89 and 179 $\\mu$m of the whole\nregion and an APEX map at 350 {\\mu}m of the IRDC. New spectral HIFI maps toward\nthe IRDC region covering low-energy H2O lines at 987 and 1113 GHz are also\npresented and combined with HIFI pointed observations of the G327 hot core. We\ninfer the physical properties of the gas through optical depth analysis and\nradiative transfer modeling. The continuum emission at 89 and 179 {\\mu}m\nfollows the thermal continuum emission at longer wavelengths, with a peak at\nthe position of the hot core, a secondary peak in the Hii region, and an\narch-like layer of hot gas west of the Hii region. The same morphology is\nobserved in the 1113 GHz line, in absorption toward all dust condensations.\nOptical depths of ~80 and 15 are estimated and correspond to column densities\nof 10^15 and 2 10^14 cm-2, for the hot core and IRDC position. These values\nindicate an H2O to H2 ratio of 3 10^-8 toward the hot core; the abundance of\nH2O does not change along the IRDC with values of some 10^-8. Infall (over ~\n20\") is detected toward the hot core position with a rate of 1-1.3 10^-2 M_sun\n/yr, high enough to overcome the radiation pressure due to the stellar\nluminosity. The source structure of the hot core region is complex, with a cold\nouter gas envelope in expansion, situated between the outflow and the observer,\nextending over 0.32 pc. The outflow is seen face-on and centered away from the\nhot core. The distribution of H2O along the IRDC is roughly constant with an\nabundance peak in the more evolved object. These water abundances are in\nagreement with previous studies in other massive objects and chemical models.\n",
"title": "Distribution of water in the G327.3-0.6 massive star-forming region"
}
| null | null | null | null | true | null |
17959
| null |
Default
| null | null |
null |
{
"abstract": " We consider a certain quotient of a polynomial ring categorified by both the\nisomorphic Green rings of the symmetric groups and Schur algebras generated by\nthe signed Young permutation modules and mixed powers respectively. They have\nbases parametrised by pairs of partitions whose second partitions are multiples\nof the odd prime $p$ the characteristic of the underlying field. We provide an\nexplicit formula rewriting a signed Young permutation module (respectively,\nmixed power) in terms of signed Young permutation modules (respectively, mixed\npowers) labelled by those pairs of partitions. As a result, for each partition\n$\\lambda$, we discovered the number of compositions $\\delta$ such that $\\delta$\ncan be rearranged to $\\lambda$ and whose partial sums of $\\delta$ are not\ndivisible by $p$.\n",
"title": "Straightening rule for an $m'$-truncated polynomial ring"
}
| null | null | null | null | true | null |
17960
| null |
Default
| null | null |
null |
{
"abstract": " Blog is becoming an increasingly popular media for information publishing.\nBesides the main content, most of blog pages nowadays also contain noisy\ninformation such as advertisements etc. Removing these unrelated elements can\nimproves user experience, but also can better adapt the content to various\ndevices such as mobile phones. Though template-based extractors are highly\naccurate, they may incur expensive cost in that a large number of template need\nto be developed and they will fail once the template is updated. To address\nthese issues, we present a novel template-independent content extractor for\nblog pages. First, we convert a blog page into a DOM-Tree, where all elements\nincluding the title and body blocks in a page correspond to subtrees. Then we\nconstruct subtree candidate set for the title and the body blocks respectively,\nand extract both spatial and content features for elements contained in the\nsubtree. SVM classifiers for the title and the body blocks are trained using\nthese features. Finally, the classifiers are used to extract the main content\nfrom blog pages. We test our extractor on 2,250 blog pages crawled from nine\nblog sites with obviously different styles and templates. Experimental results\nverify the effectiveness of our extractor.\n",
"title": "Effective Blog Pages Extractor for Better UGC Accessing"
}
| null | null | null | null | true | null |
17961
| null |
Default
| null | null |
null |
{
"abstract": " Meaningful laws of nature must be independent of the units employed to\nmeasure the variables. The principle of similitude (Rayleigh 1915) or\ndimensional homogeneity, states that only commensurable quantities (ones having\nthe same dimension) may be compared, therefore, meaningful laws of nature must\nbe homogeneous equations in their various units of measurement, a result which\nwas formalized in the $\\rm \\Pi$ theorem (Vaschy 1892; Buckingham 1914).\nHowever, most relations in allometry do not satisfy this basic requirement,\nincluding the `3/4 Law' (Kleiber 1932) that relates the basal metabolic rate\nand body mass, which it is sometimes claimed to be the most fundamental\nbiological rate (Brown et al. 2004) and the closest to a law in life sciences\n(West \\& Brown 2004). Using the $\\rm \\Pi$ theorem, here we show that it is\npossible to construct a unique homogeneous equation for the metabolic rates, in\nagreement with data in the literature. We find that the variations in the\ndependence of the metabolic rates on body mass are secondary, coming from\nvariations in the allometric dependence of the heart frequencies. This includes\nnot only different classes of animals (mammals, birds, invertebrates) but also\ndifferent exercise conditions (basal and maximal). Our results demonstrate that\nmost of the differences found in the allometric exponents (White et al. 2007)\nare due to compare incommensurable quantities and that our dimensionally\nhomogenous formula, unify these differences into a single formulation. We\ndiscuss the ecological implications of this new formulation in the context of\nthe Malthusian's, Fenchel's and the total energy consumed in a lifespan\nrelations.\n",
"title": "The Principle of Similitude in Biology: From Allometry to the Formulation of Dimensionally Homogenous `Laws'"
}
| null | null |
[
"Physics"
] | null | true | null |
17962
| null |
Validated
| null | null |
null |
{
"abstract": " Characterization of the primary events involved in the $cis-trans$\nphotoisomerization of the rhodopsin retinal chromophore was approximated by a\nminimum one-dimensional quantum-classical model. The developed mathematical\nmodel is identical to that obtained using conventional quantum-classical\napproaches, and multiparametric quantum-chemical or molecular dynamics (MD)\ncomputations were not required. The quantum subsystem of the model includes\nthree electronic states for rhodopsin: (i) the ground state, (ii) the excited\nstate, and (iii) the primary photoproduct in the ground state. The resultant\nmodel is in perfect agreement with experimental data in terms of the quantum\nyield, the time required to reach the conical intersection and to complete the\nquantum evolution, the range of the characteristic low frequencies active\nwithin the primary events of the $11-cis$ retinal isomerization, and the\ncoherent character of the photoreaction. An effective redistribution of excess\nenergy between the vibration modes of rhodopsin was revealed by analysis of the\ndissipation process. The results confirm the validity of the minimal model,\ndespite its one-dimensional character. The fundamental nature of the\nphotoreaction was therefore demonstrated using a minimum mathematical model for\nthe first time.\n",
"title": "New insight into the dynamics of rhodopsin photoisomerization from one-dimensional quantum-classical modeling"
}
| null | null | null | null | true | null |
17963
| null |
Default
| null | null |
null |
{
"abstract": " We consider implementations of high-order finite difference Weighted\nEssentially Non-Oscillatory (WENO) schemes for the Euler equations in\ncylindrical and spherical coordinate systems with radial dependence only. The\nmain concern of this work lies in ensuring both high-order accuracy and\nconservation. Three different spatial discretizations are assessed: one that is\nshown to be high-order accurate but not conservative, one conservative but not\nhigh-order accurate, and a new approach that is both high-order accurate and\nconservative. For cylindrical and spherical coordinates, we present convergence\nresults for the advection equation and the Euler equations with an acoustics\nproblem; we then use the Sod shock tube and the Sedov point-blast problems in\ncylindrical coordinates to verify our analysis and implementations.\n",
"title": "High-order schemes for the Euler equations in cylindrical/spherical coordinates"
}
| null | null | null | null | true | null |
17964
| null |
Default
| null | null |
null |
{
"abstract": " For lattice Monte Carlo simulations parallelization is crucial to make\nstudies of large systems and long simulation time feasible, while sequential\nsimulations remain the gold-standard for correlation-free dynamics. Here,\nvarious domain decomposition schemes are compared, concluding with one which\ndelivers virtually correlation-free simulations on GPU Extensive simulations of\nthe octahedron model for $2+1$ dimensional Karda--Parisi--Zhang surface growth,\nwhich is very sensitive to correlation in the site-selection dynamics, were\nperformed to show self-consistency of the parallel runs and agreement with the\nsequential algorithm. We present a GPU implementation providing a speedup of\nabout $30\\times$ over a parallel CPU implementation on a single socket and at\nleast $180\\times$ with respect to the sequential reference.\n",
"title": "Suppressing correlations in massively parallel simulations of lattice models"
}
| null | null | null | null | true | null |
17965
| null |
Default
| null | null |
null |
{
"abstract": " We prove upper and lower bounds on the effective content and logical strength\nfor a variety of natural restrictions of Hindman's Finite Sums Theorem. For\nexample, we show that Hindman's Theorem for sums of length at most 2 and 4\ncolors implies $\\mathsf{ACA}_0$. An emerging {\\em leitmotiv} is that the known\nlower bounds for Hindman's Theorem and for its restriction to sums of at most 2\nelements are already valid for a number of restricted versions which have\nsimple proofs and better computability- and proof-theoretic upper bounds than\nthe known upper bound for the full version of the theorem. We highlight the\nrole of a sparsity-like condition on the solution set, which we call apartness.\n",
"title": "New bounds on the strength of some restrictions of Hindman's Theorem"
}
| null | null | null | null | true | null |
17966
| null |
Default
| null | null |
null |
{
"abstract": " 3D Morphable Models (3DMMs) are powerful statistical models of 3D facial\nshape and texture, and among the state-of-the-art methods for reconstructing\nfacial shape from single images. With the advent of new 3D sensors, many 3D\nfacial datasets have been collected containing both neutral as well as\nexpressive faces. However, all datasets are captured under controlled\nconditions. Thus, even though powerful 3D facial shape models can be learnt\nfrom such data, it is difficult to build statistical texture models that are\nsufficient to reconstruct faces captured in unconstrained conditions\n(\"in-the-wild\"). In this paper, we propose the first, to the best of our\nknowledge, \"in-the-wild\" 3DMM by combining a powerful statistical model of\nfacial shape, which describes both identity and expression, with an\n\"in-the-wild\" texture model. We show that the employment of such an\n\"in-the-wild\" texture model greatly simplifies the fitting procedure, because\nthere is no need to optimize with regards to the illumination parameters.\nFurthermore, we propose a new fast algorithm for fitting the 3DMM in arbitrary\nimages. Finally, we have captured the first 3D facial database with relatively\nunconstrained conditions and report quantitative evaluations with\nstate-of-the-art performance. Complementary qualitative reconstruction results\nare demonstrated on standard \"in-the-wild\" facial databases. An open source\nimplementation of our technique is released as part of the Menpo Project.\n",
"title": "3D Face Morphable Models \"In-the-Wild\""
}
| null | null | null | null | true | null |
17967
| null |
Default
| null | null |
null |
{
"abstract": " In medicine, visualizing chromosomes is important for medical diagnostics,\ndrug development, and biomedical research. Unfortunately, chromosomes often\noverlap and it is necessary to identify and distinguish between the overlapping\nchromosomes. A segmentation solution that is fast and automated will enable\nscaling of cost effective medicine and biomedical research. We apply neural\nnetwork-based image segmentation to the problem of distinguishing between\npartially overlapping DNA chromosomes. A convolutional neural network is\ncustomized for this problem. The results achieved intersection over union (IOU)\nscores of 94.7% for the overlapping region and 88-94% on the non-overlapping\nchromosome regions.\n",
"title": "Image Segmentation to Distinguish Between Overlapping Human Chromosomes"
}
| null | null | null | null | true | null |
17968
| null |
Default
| null | null |
null |
{
"abstract": " Based on 13 agile transformation cases over 15 years, this article identifies\nnine challenges associated with implementing SAFe, Scrum-at-Scale, Spotify,\nLeSS, Nexus, and other mixed or customised large-scale agile frameworks. These\nchallenges should be considered by organizations aspiring to pursue a\nlarge-scale agile strategy. This article also provides recommendations for\npractitioners and agile researchers.\n",
"title": "Implementing Large-Scale Agile Frameworks: Challenges and Recommendations"
}
| null | null | null | null | true | null |
17969
| null |
Default
| null | null |
null |
{
"abstract": " A symbolic-computational algorithm, fully implemented in Maple, is described,\nthat computes explicit expressions for generating functions that enable the\nefficient computations of the expectation, variance, and higher moments, of the\nrandom variable `sum of distances to the root', defined on any given family of\nrooted ordered trees (defined by degree restrictions). Taking limits, we\nconfirm, via elementary methods, the fact, due to David Aldous, and expanded by\nSvante Janson and others, that the limiting (scaled) distributions are all the\nsame, and coincide with the limiting distribution of the same random variable,\nwhen it is defined on labeled rooted trees.\n",
"title": "On The Limiting Distributions of the Total Height On Families of Trees"
}
| null | null | null | null | true | null |
17970
| null |
Default
| null | null |
null |
{
"abstract": " HESS J1826$-$130 is an unidentified hard spectrum source discovered by\nH.E.S.S. along the Galactic plane, the spectral index being $\\Gamma$ = 1.6 with\nan exponential cut-off at about 12 TeV. While the source does not have a clear\ncounterpart at longer wavelengths, the very hard spectrum emission at TeV\nenergies implies that electrons or protons accelerated up to several hundreds\nof TeV are responsible for the emission. In the hadronic case, the VHE emission\ncan be produced by runaway cosmic-rays colliding with the dense molecular\nclouds spatially coincident with the H.E.S.S. source.\n",
"title": "HESS J1826$-$130: A Very Hard $γ$-Ray Spectrum Source in the Galactic Plane"
}
| null | null | null | null | true | null |
17971
| null |
Default
| null | null |
null |
{
"abstract": " We investigate how next-generation laser pulses at 10 PW $-$ 200 PW interact\nwith a solid target in the presence of a relativistically underdense preplasma\nproduced by amplified spontaneous emission (ASE). Laser hole boring and\nrelativistic transparency are strongly restrained due to the generation of\nelectron-positron pairs and $\\gamma$-ray photons via quantum electrodynamics\n(QED) processes. A pair plasma with a density above the initial preplasma\ndensity is formed, counteracting the electron-free channel produced by the hole\nboring. This pair-dominated plasma can block the laser transport and trigger an\navalanche-like QED cascade, efficiently transfering the laser energy to\nphotons. This renders a 1-$\\rm\\mu m$-scalelength, underdense preplasma\ncompletely opaque to laser pulses at this power level. The QED-induced opacity\ntherefore sets much higher contrast requirements for such pulse in solid-target\nexperiments than expected by classical plasma physics. Our simulations show for\nexample, that proton acceleration from the rear of a solid with a preplasma\nwould be strongly impaired.\n",
"title": "Laser opacity in underdense preplasma of solid targets due to quantum electrodynamics effects"
}
| null | null | null | null | true | null |
17972
| null |
Default
| null | null |
null |
{
"abstract": " Social dilemmas, where mutual cooperation can lead to high payoffs but\nparticipants face incentives to cheat, are ubiquitous in multi-agent\ninteraction. We wish to construct agents that cooperate with pure cooperators,\navoid exploitation by pure defectors, and incentivize cooperation from the\nrest. However, often the actions taken by a partner are (partially) unobserved\nor the consequences of individual actions are hard to predict. We show that in\na large class of games good strategies can be constructed by conditioning one's\nbehavior solely on outcomes (ie. one's past rewards). We call this\nconsequentialist conditional cooperation. We show how to construct such\nstrategies using deep reinforcement learning techniques and demonstrate, both\nanalytically and experimentally, that they are effective in social dilemmas\nbeyond simple matrix games. We also show the limitations of relying purely on\nconsequences and discuss the need for understanding both the consequences of\nand the intentions behind an action.\n",
"title": "Consequentialist conditional cooperation in social dilemmas with imperfect information"
}
| null | null | null | null | true | null |
17973
| null |
Default
| null | null |
null |
{
"abstract": " The Casimir free energy of dielectric films, both free-standing in vacuum and\ndeposited on metallic or dielectric plates, is investigated. It is shown that\nthe values of the free energy depend considerably on whether the calculation\napproach used neglects or takes into account the dc conductivity of film\nmaterial. We demonstrate that there are the material-dependent and universal\nclassical limits in the former and latter cases, respectively. The analytic\nbehavior of the Casimir free energy and entropy for a free-standing dielectric\nfilm at low temperature in found. According to our results, the Casimir entropy\ngoes to zero when the temperature vanishes if the calculation approach with\nneglected dc conductivity of a film is employed. If the dc conductivity is\ntaken into account, the Casimir entropy takes the positive value at zero\ntemperature, depending on the parameters of a film, i.e., the Nernst heat\ntheorem is violated. By considering the Casimir free energy of silica and\nsapphire films deposited on a Au plate in the framework of two calculation\napproaches, we argue that physically correct values are obtained by\ndisregarding the role of dc conductivity. A comparison with the well known\nresults for the configuration of two parallel plates is made. Finally, we\ncompute the Casimir free energy of silica, sapphire and Ge films deposited on\nhigh-resistivity Si plates of different thicknesses and demonstrate that it can\nbe positive, negative and equal to zero. Possible applications of the obtained\nresults to thin films used in microelectronics are discussed.\n",
"title": "Casimir free energy of dielectric films: Classical limit, low-temperature behavior and control"
}
| null | null | null | null | true | null |
17974
| null |
Default
| null | null |
null |
{
"abstract": " Pharmaco-epidemiology (PE) is the study of uses and effects of drugs in well\ndefined populations. As medico-administrative databases cover a large part of\nthe population, they have become very interesting to carry PE studies. Such\ndatabases provide longitudinal care pathways in real condition containing\ntimestamped care events, especially drug deliveries. Temporal pattern mining\nbecomes a strategic choice to gain valuable insights about drug uses. In this\npaper we propose DCM, a new discriminant temporal pattern mining algorithm. It\nextracts chronicle patterns that occur more in a studied population than in a\ncontrol population. We present results on the identification of possible\nassociations between hospitalizations for seizure and anti-epileptic drug\nswitches in care pathway of epileptic patients.\n",
"title": "Discriminant chronicles mining: Application to care pathways analytics"
}
| null | null | null | null | true | null |
17975
| null |
Default
| null | null |
null |
{
"abstract": " A new form of variational autoencoder (VAE) is developed, in which the joint\ndistribution of data and codes is considered in two (symmetric) forms: ($i$)\nfrom observed data fed through the encoder to yield codes, and ($ii$) from\nlatent codes drawn from a simple prior and propagated through the decoder to\nmanifest data. Lower bounds are learned for marginal log-likelihood fits\nobserved data and latent codes. When learning with the variational bound, one\nseeks to minimize the symmetric Kullback-Leibler divergence of joint density\nfunctions from ($i$) and ($ii$), while simultaneously seeking to maximize the\ntwo marginal log-likelihoods. To facilitate learning, a new form of adversarial\ntraining is developed. An extensive set of experiments is performed, in which\nwe demonstrate state-of-the-art data reconstruction and generation on several\nimage benchmark datasets.\n",
"title": "Adversarial Symmetric Variational Autoencoder"
}
| null | null | null | null | true | null |
17976
| null |
Default
| null | null |
null |
{
"abstract": " We prove that there exist hypersurfaces that contain a given closed subscheme\n$Z$ of the projective space over a finite field and intersect a given smooth\nscheme $X$ off of $Z$ smoothly, if the intersection $V = Z \\cap X$ is smooth.\nFurthermore, we can give a bound on the dimension of the singular locus of the\nhypersurface section and prescribe finitely many local conditions on the\nhypersurface. This is an analogue of a Bertini theorem of Bloch over finite\nfields and is proved using Poonen's closed point sieve. We also show a similar\ntheorem for the case where $V$ is not smooth.\n",
"title": "The singular locus of hypersurface sections containing a closed subscheme over finite fields"
}
| null | null |
[
"Mathematics"
] | null | true | null |
17977
| null |
Validated
| null | null |
null |
{
"abstract": " Inverse problems, where in broad sense the task is to learn from the noisy\nresponse about some unknown function, usually represented as the argument of\nsome known functional form, has received wide attention in the general\nscientific disciplines. How- ever, in mainstream statistics such inverse\nproblem paradigm does not seem to be as popular. In this article we provide a\nbrief overview of such problems from a statistical, particularly Bayesian,\nperspective.\nWe also compare and contrast the above class of problems with the perhaps\nmore statistically familiar inverse regression problems, arguing that this\nclass of problems contains the traditional class of inverse problems. In course\nof our review we point out that the statistical literature is very scarce with\nrespect to both the inverse paradigms, and substantial research work is still\nnecessary to develop the fields.\n",
"title": "A Statistical Perspective on Inverse and Inverse Regression Problems"
}
| null | null | null | null | true | null |
17978
| null |
Default
| null | null |
null |
{
"abstract": " This work presents a model reduction approach for problems with coherent\nstructures that propagate over time such as convection-dominated flows and\nwave-type phenomena. Traditional model reduction methods have difficulties with\nthese transport-dominated problems because propagating coherent structures\ntypically introduce high-dimensional features that require high-dimensional\napproximation spaces. The approach proposed in this work exploits the locality\nin space and time of propagating coherent structures to derive efficient\nreduced models. First, full-model solutions are approximated locally in time\nvia local reduced spaces that are adapted with basis updates during time\nstepping. The basis updates are derived from querying the full model at a few\nselected spatial coordinates. Second, the locality in space of the coherent\nstructures is exploited via an adaptive sampling scheme that selects at which\ncomponents to query the full model for computing the basis updates. Our\nanalysis shows that, in probability, the more local the coherent structure is\nin space, the fewer full-model samples are required to adapt the reduced basis\nwith the proposed adaptive sampling scheme. Numerical results on benchmark\nexamples with interacting wave-type structures and time-varying transport\nspeeds and on a model combustor of a single-element rocket engine demonstrate\nthe wide applicability of our approach and the significant runtime speedups\ncompared to full models and traditional reduced models.\n",
"title": "Model reduction for transport-dominated problems via online adaptive bases and adaptive sampling"
}
| null | null |
[
"Computer Science"
] | null | true | null |
17979
| null |
Validated
| null | null |
null |
{
"abstract": " Stochastic gradient descent (SGD) is a popular stochastic optimization method\nin machine learning. Traditional parallel SGD algorithms, e.g., SimuParallel\nSGD, often require all nodes to have the same performance or to consume equal\nquantities of data. However, these requirements are difficult to satisfy when\nthe parallel SGD algorithms run in a heterogeneous computing environment;\nlow-performance nodes will exert a negative influence on the final result. In\nthis paper, we propose an algorithm called weighted parallel SGD (WP-SGD).\nWP-SGD combines weighted model parameters from different nodes in the system to\nproduce the final output. WP-SGD makes use of the reduction in standard\ndeviation to compensate for the loss from the inconsistency in performance of\nnodes in the cluster, which means that WP-SGD does not require that all nodes\nconsume equal quantities of data. We also analyze the theoretical feasibility\nof running two other parallel SGD algorithms combined with WP-SGD in a\nheterogeneous environment. The experimental results show that WP-SGD\nsignificantly outperforms the traditional parallel SGD algorithms on\ndistributed training systems with an unbalanced workload.\n",
"title": "Weighted parallel SGD for distributed unbalanced-workload training system"
}
| null | null | null | null | true | null |
17980
| null |
Default
| null | null |
null |
{
"abstract": " Precise trajectory control near ground is difficult for multi-rotor drones,\ndue to the complex ground effects caused by interactions between multi-rotor\nairflow and the environment. Conventional control methods often fail to\nproperly account for these complex effects and fall short in accomplishing\nsmooth landing. In this paper, we present a novel deep-learning-based robust\nnonlinear controller (Neural-Lander) that improves control performance of a\nquadrotor during landing. Our approach blends together a nominal dynamics model\ncoupled with a Deep Neural Network (DNN) that learns the high-order\ninteractions. We employ a novel application of spectral normalization to\nconstrain the DNN to have bounded Lipschitz behavior. Leveraging this Lipschitz\nproperty, we design a nonlinear feedback linearization controller using the\nlearned model and prove system stability with disturbance rejection. To the\nbest of our knowledge, this is the first DNN-based nonlinear feedback\ncontroller with stability guarantees that can utilize arbitrarily large neural\nnets. Experimental results demonstrate that the proposed controller\nsignificantly outperforms a baseline linear proportional-derivative (PD)\ncontroller in both 1D and 3D landing cases. In particular, we show that\ncompared to the PD controller, Neural-Lander can decrease error in z direction\nfrom 0.13m to zero, and mitigate average x and y drifts by 90% and 34%\nrespectively, in 1D landing. Meanwhile, Neural-Lander can decrease z error from\n0.12m to zero, in 3D landing. We also empirically show that the DNN generalizes\nwell to new test inputs outside the training domain.\n",
"title": "Neural Lander: Stable Drone Landing Control using Learned Dynamics"
}
| null | null | null | null | true | null |
17981
| null |
Default
| null | null |
null |
{
"abstract": " This communication presents a longitudinal model-free control approach for\ncomputing the wheel torque command to be applied on a vehicle. This setting\nenables us to overcome the problem of unknown vehicle parameters for generating\na suitable control law. An important parameter in this control setting is made\ntime-varying for ensuring finite-time stability. Several convincing computer\nsimulations are displayed and discussed. Overshoots become therefore smaller.\nThe driving comfort is increased and the robustness to time-delays is improved.\n",
"title": "Finite-Time Stabilization of Longitudinal Control for Autonomous Vehicles via a Model-Free Approach"
}
| null | null | null | null | true | null |
17982
| null |
Default
| null | null |
null |
{
"abstract": " This new book by cosmologists Geraint F. Lewis and Luke A. Barnes is another\nentry in the long list of cosmology-centered physics books intended for a large\naudience. While many such books aim at advancing a novel scientific theory, A\nFortunate Universe has no such scientific pretense. Its goals are to assert\nthat the universe is fine-tuned for life, to defend that this fact can\nreasonably motivate further scientific inquiry as to why it is so, and to show\nthat the multiverse and intelligent design hypotheses are reasonable proposals\nto explain this fine-tuning. This book's potential contribution, therefore,\nlies in how convincingly and efficiently it can make that case.\n",
"title": "Review of Geraint F. Lewis and Luke A. Barnes, A Fortunate Universe: Life in a Finely Tuned Cosmos"
}
| null | null |
[
"Physics"
] | null | true | null |
17983
| null |
Validated
| null | null |
null |
{
"abstract": " A novel sparsity-based algorithm for audio inpainting is proposed by\ntranslating the SPADE algorithm by Kitić et. al.---the state-of-the-art for\naudio declipping---into the task of audio inpainting. SPAIN (SParse Audio\nINpainter) comes in synthesis and analysis variants. Experiments show that both\nA-SPAIN and S-SPAIN outperform other sparsity-based inpainting algorithms and\nthat A-SPAIN performs on a par with the state-of-the-art method based on linear\nprediction.\n",
"title": "Introducing SPAIN (SParse Audion INpainter)"
}
| null | null | null | null | true | null |
17984
| null |
Default
| null | null |
null |
{
"abstract": " Identifying palindromes in sequences has been an interesting line of research\nin combinatorics on words and also in computational biology, after the\ndiscovery of the relation of palindromes in the DNA sequence with the HIV\nvirus. Efficient algorithms for the factorization of sequences into palindromes\nand maximal palindromes have been devised in recent years. We extend these\nstudies by allowing gaps in decompositions and errors in palindromes, and also\nimposing a lower bound to the length of acceptable palindromes.\nWe first present an algorithm for obtaining a palindromic decomposition of a\nstring of length n with the minimal total gap length in time O(n log n * g) and\nspace O(n g), where g is the number of allowed gaps in the decomposition. We\nthen consider a decomposition of the string in maximal \\delta-palindromes (i.e.\npalindromes with \\delta errors under the edit or Hamming distance) and g\nallowed gaps. We present an algorithm to obtain such a decomposition with the\nminimal total gap length in time O(n (g + \\delta)) and space O(n g).\n",
"title": "Palindromic Decompositions with Gaps and Errors"
}
| null | null | null | null | true | null |
17985
| null |
Default
| null | null |
null |
{
"abstract": " Remanufacturing is a significant factor in securing sustainability through a\ncircular economy. Sorting plays a significant role in remanufacturing\npre-processing inspections. Its significance can increase when remanufacturing\nfacilities encounter extreme situations, such as abnormally huge core arrivals.\nOur main objective in this work is switching from less efficient to a more\nefficient model and to characterize extreme behavior of core arrival in\nremanufacturing and applying the developed model to triage cores. Central\ntendency core flow models are not sufficient to handle extreme situations,\nhowever, complementary Extreme Value (EV) approaches have shown to improve\nmodel efficiency. Extreme core flows to remanufacturing facilities are rare but\nstill likely and can adversely affect remanufacturing business operations. In\nthis investigation, extreme end-of-use core flow is modelled by a threshold\napproach using the Generalized Pareto Distribution (GPD). It is shown that GPD\nhas better performance than its maxima-block GEV counterpart from practical and\ndata efficiency perspectives. The model is validated by a synthesized big\ndataset, tested by sophisticated statistical Anderson Darling (AD) test, and is\napplied to a case of extreme flow to a valve shop in order to predict\nprobability of over-capacity arrivals that is critical in remanufacturing\nbusiness management. Finally, the GPD model combined with triage strategies is\nused to initiate investigations into the efficacy of different triage methods\nin remanufacturing operations.\n",
"title": "End-of-Use Core Triage in Extreme Scenarios Based on a Threshold Approach"
}
| null | null | null | null | true | null |
17986
| null |
Default
| null | null |
null |
{
"abstract": " A temperature (T)-dependent coarse-grained (CG) Hamiltonian of polyethylene\nglycol/oxide (PEG/PEO) in aqueous solution is reported to be used in\nimplicit-solvent material models in a wide temperature (i.e., solvent quality)\nrange. The T-dependent nonbonded CG interactions are derived from a combined\n\"bottom-up\" and \"top-down\" approach. The pair potentials calculated from\natomistic replica-exchange molecular dynamics simulations in combination with\nthe iterative Boltzmann inversion are post-refined by benchmarking to\nexperimental data of the radius of gyration. For better handling and a fully\ncontinuous transferability in T-space, the pair potentials are conveniently\ntruncated and mapped to an analytic formula with three structural parameters\nexpressed as explicit continuous functions of T. It is then demonstrated that\nthis model without further adjustments successfully reproduces other\nexperimentally known key thermodynamic properties of semi-dilute PEG solutions\nsuch as the full equation of state (i.e., T-dependent osmotic pressure) for\nvarious chain lengths as well as their cloud point (or collapse) temperature.\n",
"title": "A temperature-dependent implicit-solvent model of polyethylene glycol in aqueous solution"
}
| null | null | null | null | true | null |
17987
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we present libDirectional, a MATLAB library for directional\nstatistics and directional estimation. It supports a variety of commonly used\ndistributions on the unit circle, such as the von Mises, wrapped normal, and\nwrapped Cauchy distributions. Furthermore, various distributions on\nhigher-dimensional manifolds such as the unit hypersphere and the hypertorus\nare available. Based on these distributions, several recursive filtering\nalgorithms in libDirectional allow estimation on these manifolds. The\nfunctionality is implemented in a clear, well-documented, and object-oriented\nstructure that is both easy to use and easy to extend.\n",
"title": "Directional Statistics and Filtering Using libDirectional"
}
| null | null | null | null | true | null |
17988
| null |
Default
| null | null |
null |
{
"abstract": " A novel idea is proposed for a natural solution of the dark energy and its\ncosmic coincidence problem. The existence of local antigravity sources,\nassociated with astrophysical matter configurations distributed throughout the\nuniverse, can lead to a recent cosmic acceleration effect. Various physical\ntheories can be compatible with this idea, but here, in order to test our\nproposal, we focus on quantum originated spherically symmetric metrics matched\nwith the cosmological evolution through the simplest Swiss cheese model. In the\ncontext of asymptotically safe gravity, we have explained the observed amount\nof dark energy using Newton's constant, the galaxy or cluster length scales,\nand dimensionless order one parameters predicted by the theory, without\nfine-tuning or extra unproven energy scales. The interior modified\nSchwarzschild-de Sitter metric allows us to approximately interpret this result\nas that the standard cosmological constant is a composite quantity made of the\nabove parameters, instead of a fundamental one.\n",
"title": "A solution of the dark energy and its coincidence problem based on local antigravity sources without fine-tuning or new scales"
}
| null | null | null | null | true | null |
17989
| null |
Default
| null | null |
null |
{
"abstract": " Self-paced learning and hard example mining re-weight training instances to\nimprove learning accuracy. This paper presents two improved alternatives based\non lightweight estimates of sample uncertainty in stochastic gradient descent\n(SGD): the variance in predicted probability of the correct class across\niterations of mini-batch SGD, and the proximity of the correct class\nprobability to the decision threshold. Extensive experimental results on six\ndatasets show that our methods reliably improve accuracy in various network\narchitectures, including additional gains on top of other popular training\ntechniques, such as residual learning, momentum, ADAM, batch normalization,\ndropout, and distillation.\n",
"title": "Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples"
}
| null | null | null | null | true | null |
17990
| null |
Default
| null | null |
null |
{
"abstract": " To investigate the electronic structure of Weyl semimetals Ta$Pn$ ($Pn=$P,\nAs), optical conductivity [$\\sigma(\\omega)$] spectra are measured over a wide\nrange of photon energies and temperatures, and these measured values are\ncompared with band calculations. Two significant structures can be observed: a\nbending structure at $\\hbar\\omega\\sim$85 meV in TaAs, and peaks at\n$\\hbar\\omega\\sim$ 50 meV (TaP) and $\\sim$30 meV (TaAs). The bending structure\ncan be explained by the interband transition between saddle points connecting a\nset of $W_2$ Weyl points. The temperature dependence of the peak intensity can\nbe fitted by assuming the interband transition between saddle points connecting\na set of $W_1$ Weyl points. Owing to the different temperature dependence of\nthe Drude weight in both materials, it is found that the Weyl points of TaAs\nare located near the Fermi level, whereas those of TaP are further away.\n",
"title": "Optical signature of Weyl electronic structures in tantalum pnictides Ta$Pn$ ($Pn=$ P, As)"
}
| null | null | null | null | true | null |
17991
| null |
Default
| null | null |
null |
{
"abstract": " The purpose of this study is to analyze cyber security and security practices\nof electronic information and network system, network threats, and techniques\nto prevent the cyber attacks in hotels. Helping the information technology\ndirectors and chief information officers (CIO) is the aim of this study to\nadvance policy for security of electronic information in hotels and suggesting\nsome techniques and tools to secure the computer networks. This research is\ncompletely qualitative while the case study and interviews have done in 5\nrandom hotels in Reno, Nevada, United States of America. The interview has done\nwith 50 hotel guests, 10 front desk employees, 3 IT manager and 2 assistant of\nGeneral manager. The results show that hotels' cyber security is very low and\nhotels are very vulnerable in this regard and at the end, the implications and\ncontribution of the study is mentioned.\n",
"title": "A study of cyber security in hospitality industry- threats and countermeasures: case study in Reno, Nevada"
}
| null | null | null | null | true | null |
17992
| null |
Default
| null | null |
null |
{
"abstract": " We develop the theory of weak Fraisse categories, where the crucial concept\nis the weak amalgamation property, discovered relatively recently in model\ntheory. We show that, in a suitable framework, every weak Fraisse category has\nits unique limit, a special object in a bigger category, characterized by\ncertain variant of injectivity. This significantly extends the known theory of\nFraisse limits.\n",
"title": "Weak Fraisse categories"
}
| null | null | null | null | true | null |
17993
| null |
Default
| null | null |
null |
{
"abstract": " Music recommendation services collectively spin billions of songs for\nmillions of listeners on a daily basis. Users can typically listen to a variety\nof songs tailored to their personal tastes and preferences. Music is not the\nonly type of content encountered in these services, however. Advertisements are\ngenerally interspersed throughout the music stream to generate revenue for the\nbusiness. Additional content may include artist messaging, ticketing, sports,\nnews and weather. In this paper, we discuss issues that arise when multiple\ncontent providers are stakeholders in the recommendation process. These\nstakeholders each have their own objectives and must work in concert to sustain\na healthy music recommendation service.\n",
"title": "Multiple Stakeholders in Music Recommender Systems"
}
| null | null | null | null | true | null |
17994
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the large time behavior of solution for a class of\nnonlinear massless Dirac equations in $R^{1+1}$. It is shown that the solution\nwill tend to travelling wave solution when time tends to infinity.\n",
"title": "Large time behavior of solution to nonlinear Dirac equation in $1+1$ dimensions"
}
| null | null | null | null | true | null |
17995
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies mathematical properties of reaction systems that was\nintroduced by Enrenfeucht and Rozenberg as computational models inspired by\nbiochemical reaction in the living cells. In particular, we continue the study\non the generative power of functions specified by minimal reaction systems\nunder composition initiated by Salomaa. Allowing degenerate reaction systems,\nfunctions specified by minimal reaction systems over a quarternary alphabet\nthat are permutations generate the alternating group on the power set of the\nbackground set.\n",
"title": "Compositions of Functions and Permutations Specified by Minimal Reaction Systems"
}
| null | null | null | null | true | null |
17996
| null |
Default
| null | null |
null |
{
"abstract": " Chemical reaction networks with generalized mass-action kinetics lead to\npower-law dynamical systems. As a simple example, we consider the Lotka\nreactions and the resulting planar ODE. We characterize the parameters\n(positive coefficients and real exponents) for which the unique positive\nequilibrium is a center.\n",
"title": "The center problem for the Lotka reactions with generalized mass-action kinetics"
}
| null | null | null | null | true | null |
17997
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the missing sample recovery problem using methods\nbased on sparse approximation. In this regard, we investigate the algorithms\nused for solving the inverse problem associated with the restoration of missed\nsamples of image signal. This problem is also known as inpainting in the\ncontext of image processing and for this purpose, we suggest an iterative\nsparse recovery algorithm based on constrained $l_1$-norm minimization with a\nnew fidelity metric. The proposed metric called Convex SIMilarity (CSIM) index,\nis a simplified version of the Structural SIMilarity (SSIM) index, which is\nconvex and error-sensitive. The optimization problem incorporating this\ncriterion, is then solved via Alternating Direction Method of Multipliers\n(ADMM). Simulation results show the efficiency of the proposed method for\nmissing sample recovery of 1D patch vectors and inpainting of 2D image signals.\n",
"title": "Recovery of Missing Samples Using Sparse Approximation via a Convex Similarity Measure"
}
| null | null | null | null | true | null |
17998
| null |
Default
| null | null |
null |
{
"abstract": " As one kind of skin cancer, melanoma is very dangerous. Dermoscopy based\nearly detection and recarbonization strategy is critical for melanoma therapy.\nHowever, well-trained dermatologists dominant the diagnostic accuracy. In order\nto solve this problem, many effort focus on developing automatic image analysis\nsystems. Here we report a novel strategy based on deep learning technique, and\nachieve very high skin lesion segmentation and melanoma diagnosis accuracy: 1)\nwe build a segmentation neural network (skin_segnn), which achieved very high\nlesion boundary detection accuracy; 2) We build another very deep neural\nnetwork based on Google inception v3 network (skin_recnn) and its well-trained\nweight. The novel designed transfer learning based deep neural network\nskin_inceptions_v3_nn helps to achieve a high prediction accuracy.\n",
"title": "Skin cancer reorganization and classification with deep neural network"
}
| null | null | null | null | true | null |
17999
| null |
Default
| null | null |
null |
{
"abstract": " We extend Rubio de Francia's extrapolation theorem for functions valued in\nUMD Banach function spaces, leading to short proofs of some new and known\nresults. In particular we prove Littlewood-Paley-Rubio de Francia-type\nestimates and boundedness of variational Carleson operators for Banach function\nspaces with UMD concavifications.\n",
"title": "Rescaled extrapolation for vector-valued functions"
}
| null | null | null | null | true | null |
18000
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.