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": " The adoption of the distributed paradigm has allowed applications to increase\ntheir scalability, robustness and fault tolerance, but it has also complicated\ntheir structure, leading to an exponential growth of the applications'\nconfiguration space and increased difficulty in predicting their performance.\nIn this work, we describe a novel, automated profiling methodology that makes\nno assumptions on application structure. Our approach utilizes oblique Decision\nTrees in order to recursively partition an application's configuration space in\ndisjoint regions, choose a set of representative samples from each subregion\naccording to a defined policy and return a model for the entire space as a\ncomposition of linear models over each subregion. An extensive evaluation over\nreal-life applications and synthetic performance functions showcases that our\nscheme outperforms other state-of-the-art profiling methodologies. It\nparticularly excels at reflecting abnormalities and discontinuities of the\nperformance function, allowing the user to influence the sampling policy based\non the modeling accuracy and the space coverage.\n",
"title": "A Decision Tree Based Approach Towards Adaptive Profiling of Distributed Applications"
}
| null | null | null | null | true | null |
9301
| null |
Default
| null | null |
null |
{
"abstract": " Spatial distributions of other cell interference (OCIF) and interference to\nown-cell power ratio (IOPR) with reference to the distance between a mobile and\nits serving base station (BS) are modeled for the down-link reception of\ncellular systems based on the best-cell configuration instead of the\nnearest-cell configuration. This enables a more realistic evaluation of two\ncompeting objectives in network dimensioning: coverage and rate capacity. More\noutcomes useful for dynamic network dimensioning are also derived, including\nmaximum BS transmission power per cell size and the cell density required for\nan adequate coverage of a given traffic density.\n",
"title": "A Tractable Approach to Dynamic Network Dimensioning Based on the Best-cell Configuration"
}
| null | null | null | null | true | null |
9302
| null |
Default
| null | null |
null |
{
"abstract": " In order to understand underlying processes governing environmental and\nphysical processes, and predict future outcomes, a complex computer model is\nfrequently required to simulate these dynamics. However there is inevitably\nuncertainty related to the exact parametric form or the values of such\nparameters to be used when developing these simulators, with \\emph{ranges} of\nplausible values prevalent in the literature. Systematic errors introduced by\nfailing to account for these uncertainties have the potential to have a large\neffect on resulting estimates in unknown quantities of interest. Due to the\ncomplexity of these types of models, it is often unfeasible to run large\nnumbers of training runs that are usually required for full statistical\nemulators of the environmental processes. We therefore present a method for\naccounting for uncertainties in complex environmental simulators without the\nneed for very large numbers of training runs and illustrate the method through\nan application to the Met Office's atmospheric transport model NAME. We\nconclude that there are two principle parameters that are linked with\nvariability in NAME outputs, namely the free tropospheric turbulence parameter\nand particle release height. Our results suggest the former should be\nsignificantly larger than is currently implemented as a default in NAME, whilst\nchanges in the latter most likely stem from inconsistencies between the model\nspecified ground height at the observation locations and the true height at\nthis location. Estimated discrepancies from independent data are consistent\nwith the discrepancy between modelled and true ground height.\n",
"title": "Parametric uncertainty in complex environmental models: a cheap emulation approach for models with high-dimensional output"
}
| null | null | null | null | true | null |
9303
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a method for learning the dynamics of complex nonlinear systems\nbased on deep generative models over temporal segments of states and actions.\nUnlike dynamics models that operate over individual discrete timesteps, we\nlearn the distribution over future state trajectories conditioned on past\nstate, past action, and planned future action trajectories, as well as a latent\nprior over action trajectories. Our approach is based on convolutional\nautoregressive models and variational autoencoders. It makes stable and\naccurate predictions over long horizons for complex, stochastic systems,\neffectively expressing uncertainty and modeling the effects of collisions,\nsensory noise, and action delays. The learned dynamics model and action prior\ncan be used for end-to-end, fully differentiable trajectory optimization and\nmodel-based policy optimization, which we use to evaluate the performance and\nsample-efficiency of our method.\n",
"title": "Prediction and Control with Temporal Segment Models"
}
| null | null | null | null | true | null |
9304
| null |
Default
| null | null |
null |
{
"abstract": " Artificial intelligence (AI) is intrinsically data-driven. It calls for the\napplication of statistical concepts through human-machine collaboration during\ngeneration of data, development of algorithms, and evaluation of results. This\npaper discusses how such human-machine collaboration can be approached through\nthe statistical concepts of population, question of interest,\nrepresentativeness of training data, and scrutiny of results (PQRS). The PQRS\nworkflow provides a conceptual framework for integrating statistical ideas with\nhuman input into AI products and research. These ideas include experimental\ndesign principles of randomization and local control as well as the principle\nof stability to gain reproducibility and interpretability of algorithms and\ndata results. We discuss the use of these principles in the contexts of\nself-driving cars, automated medical diagnoses, and examples from the authors'\ncollaborative research.\n",
"title": "Artificial Intelligence and Statistics"
}
| null | null | null | null | true | null |
9305
| null |
Default
| null | null |
null |
{
"abstract": " Deep stacked RNNs are usually hard to train. Adding shortcut connections\nacross different layers is a common way to ease the training of stacked\nnetworks. However, extra shortcuts make the recurrent step more complicated. To\nsimply the stacked architecture, we propose a framework called shortcut block,\nwhich is a marriage of the gating mechanism and shortcuts, while discarding the\nself-connected part in LSTM cell. We present extensive empirical experiments\nshowing that this design makes training easy and improves generalization. We\npropose various shortcut block topologies and compositions to explore its\neffectiveness. Based on this architecture, we obtain a 6% relatively\nimprovement over the state-of-the-art on CCGbank supertagging dataset. We also\nget comparable results on POS tagging task.\n",
"title": "Shortcut Sequence Tagging"
}
| null | null | null | null | true | null |
9306
| null |
Default
| null | null |
null |
{
"abstract": " By using N-body hydrodynamical cosmological simulations in which the\nchemistry of major metals and molecules is consistently solved for, we study\nthe interaction of metallic fine-structure lines with the CMB. Our analysis\nshows that the collisional induced emissions in the OI 145 $\\mu$m and CII 158\n$\\mu$m lines during reionization introduce a distortion of the CMB spectrum at\nlow frequencies ($\\nu < 300$ GHz) with amplitudes up to $\\Delta\nI_{\\nu}/B_{\\nu}(T_{\\rm CMB})\\sim 10^{-8}$-$10^{-7}$, i.e., at the $\\sim 0.1$\npercent level of FIRAS upper limits. Shorter wavelength fine-structure\ntransitions (OI 63 $\\mu$m, FeII 26 $\\mu$m, and SiII 35 $\\mu$m) typically sample\nthe reionization epoch at higher observing frequencies ($\\nu > 400$ GHz). This\ncorresponds to the Wien tail of the CMB spectrum and the distortion level\ninduced by those lines may be as high as $\\Delta I_{\\nu}/B_{\\nu}(T_{\\rm\nCMB})\\sim 10^{-4}$. The angular anisotropy produced by these lines should be\nmore relevant at higher frequencies: while practically negligible at $\\nu=145\n$GHz, signatures from CII 158 $\\mu$m and OI 145 $\\mu$m should amount to 1%-5%\nof the anisotropy power measured at $l \\sim 5000$ and $\\nu=220 $GHz by the ACT\nand SPT collaborations (after assuming $\\Delta \\nu_{\\rm obs}/\\nu_{\\rm\nobs}\\simeq 0.005$ for the line observations). Our simulations show that\nanisotropy maps from different lines (e.g., OI 145 $\\mu$m and CII 158 $\\mu$m)\nat the same redshift show a very high degree ($>0.8$) of spatial correlation,\nallowing for the use of observations at different frequencies to unveil the\nsame snapshot of the reionization epoch. Finally, our simulations demonstrate\nthat line-emission anisotropies extracted in narrow frequency/redshift shells\nare practically uncorrelated in frequency space, thus enabling standard methods\nfor removal of foregrounds that vary smoothly in frequency, just as in HI 21 cm\nstudies.\n",
"title": "Distortions of the Cosmic Microwave Background through cooling lines during the epoch of Reionization"
}
| null | null | null | null | true | null |
9307
| null |
Default
| null | null |
null |
{
"abstract": " One of the big restrictions in brain computer interface field is the very\nlimited training samples, it is difficult to build a reliable and usable system\nwith such limited data. Inspired by generative adversarial networks, we propose\na conditional Deep Convolutional Generative Adversarial (cDCGAN) Networks\nmethod to generate more artificial EEG signal automatically for data\naugmentation to improve the performance of convolutional neural networks in\nbrain computer interface field and overcome the small training dataset\nproblems. We evaluate the proposed cDCGAN method on BCI competition dataset of\nmotor imagery. The results show that the generated artificial EEG data from\nGaussian noise can learn the features from raw EEG data and has no less than\nthe classification accuracy of raw EEG data in the testing dataset. Also by\nusing generated artificial data can effectively improve classification accuracy\nat the same model with limited training data.\n",
"title": "Improving brain computer interface performance by data augmentation with conditional Deep Convolutional Generative Adversarial Networks"
}
| null | null | null | null | true | null |
9308
| null |
Default
| null | null |
null |
{
"abstract": " Our goal is to learn a semantic parser that maps natural language utterances\ninto executable programs when only indirect supervision is available: examples\nare labeled with the correct execution result, but not the program itself.\nConsequently, we must search the space of programs for those that output the\ncorrect result, while not being misled by spurious programs: incorrect programs\nthat coincidentally output the correct result. We connect two common learning\nparadigms, reinforcement learning (RL) and maximum marginal likelihood (MML),\nand then present a new learning algorithm that combines the strengths of both.\nThe new algorithm guards against spurious programs by combining the systematic\nsearch traditionally employed in MML with the randomized exploration of RL, and\nby updating parameters such that probability is spread more evenly across\nconsistent programs. We apply our learning algorithm to a new neural semantic\nparser and show significant gains over existing state-of-the-art results on a\nrecent context-dependent semantic parsing task.\n",
"title": "From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood"
}
| null | null | null | null | true | null |
9309
| null |
Default
| null | null |
null |
{
"abstract": " Object Transfiguration replaces an object in an image with another object\nfrom a second image. For example it can perform tasks like \"putting exactly\nthose eyeglasses from image A on the nose of the person in image B\". Usage of\nexemplar images allows more precise specification of desired modifications and\nimproves the diversity of conditional image generation. However, previous\nmethods that rely on feature space operations, require paired data and/or\nappearance models for training or disentangling objects from background. In\nthis work, we propose a model that can learn object transfiguration from two\nunpaired sets of images: one set containing images that \"have\" that kind of\nobject, and the other set being the opposite, with the mild constraint that the\nobjects be located approximately at the same place. For example, the training\ndata can be one set of reference face images that have eyeglasses, and another\nset of images that have not, both of which spatially aligned by face landmarks.\nDespite the weak 0/1 labels, our model can learn an \"eyeglasses\" subspace that\ncontain multiple representatives of different types of glasses. Consequently,\nwe can perform fine-grained control of generated images, like swapping the\nglasses in two images by swapping the projected components in the \"eyeglasses\"\nsubspace, to create novel images of people wearing eyeglasses.\nOverall, our deterministic generative model learns disentangled attribute\nsubspaces from weakly labeled data by adversarial training. Experiments on\nCelebA and Multi-PIE datasets validate the effectiveness of the proposed model\non real world data, in generating images with specified eyeglasses, smiling,\nhair styles, and lighting conditions etc. The code is available online.\n",
"title": "GeneGAN: Learning Object Transfiguration and Attribute Subspace from Unpaired Data"
}
| null | null | null | null | true | null |
9310
| null |
Default
| null | null |
null |
{
"abstract": " Segmenting foreground object from a video is a challenging task because of\nthe large deformations of the objects, occlusions, and background clutter. In\nthis paper, we propose a frame-by-frame but computationally efficient approach\nfor video object segmentation by clustering visually similar generic object\nsegments throughout the video. Our algorithm segments various object instances\nappearing in the video and then perform clustering in order to group visually\nsimilar segments into one cluster. Since the object that needs to be segmented\nappears in most part of the video, we can retrieve the foreground segments from\nthe cluster having maximum number of segments, thus filtering out noisy\nsegments that do not represent any object. We then apply a track and fill\napproach in order to localize the objects in the frames where the object\nsegmentation framework fails to segment any object. Our algorithm performs\ncomparably to the recent automatic methods for video object segmentation when\nbenchmarked on DAVIS dataset while being computationally much faster.\n",
"title": "Flow-free Video Object Segmentation"
}
| null | null | null | null | true | null |
9311
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we develop a novel paradigm, namely hypergraph shift, to find\nrobust graph modes by probabilistic voting strategy, which are semantically\nsound besides the self-cohesiveness requirement in forming graph modes. Unlike\nthe existing techniques to seek graph modes by shifting vertices based on\npair-wise edges (i.e, an edge with $2$ ends), our paradigm is based on shifting\nhigh-order edges (hyperedges) to deliver graph modes. Specifically, we convert\nthe problem of seeking graph modes as the problem of seeking maximizers of a\nnovel objective function with the aim to generate good graph modes based on\nsifting edges in hypergraphs. As a result, the generated graph modes based on\ndense subhypergraphs may more accurately capture the object semantics besides\nthe self-cohesiveness requirement. We also formally prove that our technique is\nalways convergent. Extensive empirical studies on synthetic and real world data\nsets are conducted on clustering and graph matching. They demonstrate that our\ntechniques significantly outperform the existing techniques.\n",
"title": "Finding Modes by Probabilistic Hypergraphs Shifting"
}
| null | null | null | null | true | null |
9312
| null |
Default
| null | null |
null |
{
"abstract": " A set of points in d-dimensional Euclidean space is almost equidistant if\namong any three points of the set, some two are at distance 1. We show that an\nalmost-equidistant set in $\\mathbb{R}^d$ has cardinality $O(d^{4/3})$.\n",
"title": "Bounding the size of an almost-equidistant set in Euclidean space"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
9313
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate the extent to which the weak equivalences in a model category\ncan be equipped with algebraic structure. We prove, for instance, that there\nexists a monad T such that a morphism of topological spaces admits T-algebra\nstructure if and only it is a weak homotopy equivalence. Likewise for\nquasi-isomorphisms and many other examples. The basic trick is to consider\ninjectivity in arrow categories. Using algebraic injectivity and cone\ninjectivity we obtain general results about the extent to which the weak\nequivalences in a combinatorial model category can be equipped with algebraic\nstructure.\n",
"title": "Equipping weak equivalences with algebraic structure"
}
| null | null | null | null | true | null |
9314
| null |
Default
| null | null |
null |
{
"abstract": " Treewidth is a parameter that measures how tree-like a relational instance\nis, and whether it can reasonably be decomposed into a tree. Many computation\ntasks are known to be tractable on databases of small treewidth, but computing\nthe treewidth of a given instance is intractable. This article is the first\nlarge-scale experimental study of treewidth and tree decompositions of\nreal-world database instances (25 datasets from 8 different domains, with sizes\nranging from a few thousand to a few million vertices). The goal is to\ndetermine which data, if any, can benefit of the wealth of algorithms for\ndatabases of small treewidth. For each dataset, we obtain upper and lower bound\nestimations of their treewidth, and study the properties of their tree\ndecompositions. We show in particular that, even when treewidth is high, using\npartial tree decompositions can result in data structures that can assist\nalgorithms.\n",
"title": "An Experimental Study of the Treewidth of Real-World Graph Data (Extended Version)"
}
| null | null | null | null | true | null |
9315
| null |
Default
| null | null |
null |
{
"abstract": " Many microbial systems are known to actively reshape their proteomes in\nresponse to changes in growth conditions induced e.g. by nutritional stress or\nantibiotics. Part of the re-allocation accounts for the fact that, as the\ngrowth rate is limited by targeting specific metabolic activities, cells simply\nrespond by fine-tuning their proteome to invest more resources into the\nlimiting activity (i.e. by synthesizing more proteins devoted to it). However,\nthis is often accompanied by an overall re-organization of metabolism, aimed at\nimproving the growth yield under limitation by re-wiring resource through\ndifferent pathways. While both effects impact proteome composition, the latter\nunderlies a more complex systemic response to stress. By focusing on E. coli's\n`acetate switch', we use mathematical modeling and a re-analysis of empirical\ndata to show that the transition from a predominantly fermentative to a\npredominantly respirative metabolism in carbon-limited growth results from the\ntrade-off between maximizing the growth yield and minimizing its costs in terms\nof required the proteome share. In particular, E. coli's metabolic phenotypes\nappear to be Pareto-optimal for these objective functions over a broad range of\ndilutions.\n",
"title": "A yield-cost tradeoff governs Escherichia coli's decision between fermentation and respiration in carbon-limited growth"
}
| null | null | null | null | true | null |
9316
| null |
Default
| null | null |
null |
{
"abstract": " Being an unsupervised machine learning and data mining technique,\nbiclustering and its multimodal extensions are becoming popular tools for\nanalysing object-attribute data in different domains. Apart from conventional\nclustering techniques, biclustering is searching for homogeneous groups of\nobjects while keeping their common description, e.g., in binary setting, their\nshared attributes. In bioinformatics, biclustering is used to find genes, which\nare active in a subset of situations, thus being candidates for biomarkers.\nHowever, the authors of those biclustering techniques that are popular in gene\nexpression analysis, may overlook the existing methods. For instance, BiMax\nalgorithm is aimed at finding biclusters, which are well-known for decades as\nformal concepts. Moreover, even if bioinformatics classify the biclustering\nmethods according to reasonable domain-driven criteria, their classification\ntaxonomies may be different from survey to survey and not full as well. So, in\nthis paper we propose to use concept lattices as a tool for taxonomy building\n(in the biclustering domain) and attribute exploration as means for\ncross-domain taxonomy completion.\n",
"title": "Towards a Unified Taxonomy of Biclustering Methods"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9317
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we consider the defocusing energy critical wave equation with a\ntrapping potential in dimension $3$. We prove that the set of initial data for\nwhich solutions scatter to an unstable excited state $(\\phi, 0)$ forms a finite\nco-dimensional path connected $C^1$ manifold in the energy space. This manifold\nis a global and unique center-stable manifold associated with $(\\phi,0)$. It is\nconstructed in a first step locally around any solution scattering to $\\phi$,\nwhich might be very far away from $\\phi$ in the $\\dot{H}^1\\times\nL^2(\\mathbb{R}^3)$ norm. In a second crucial step a no-return property is\nproved for any solution which starts near, but not on the local manifolds. This\nensures that the local manifolds form a global one. Scattering to an unstable\nsteady state is therefore a non-generic behavior, in a strong topological sense\nin the energy space. This extends our previous result [18] to the nonradial\ncase. The new ingredients here are (i) application of the reversed Strichartz\nestimate from [3] to construct a local center stable manifold near any solution\nthat scatters to $(\\phi, 0)$. This is needed since the endpoint of the standard\nStrichartz estimates fails nonradially. (ii) The nonradial channel of energy\nestimate introduced by Duyckaerts-Kenig-Merle [14], which is used to show that\nsolutions that start off but near the local manifolds associated with $\\phi$\nemit some amount of energy into the far field in excess of the amount of energy\nbeyond that of the steady state $\\phi$.\n",
"title": "Global center stable manifold for the defocusing energy critical wave equation with potential"
}
| null | null | null | null | true | null |
9318
| null |
Default
| null | null |
null |
{
"abstract": " Deep Neural Networks have impressive classification performance, but this\ncomes at the expense of significant computational resources at inference time.\nAutonomous Underwater Vehicles use low-power embedded systems for sonar image\nperception, and cannot execute large neural networks in real-time. We propose\nthe use of max-pooling aggressively, and we demonstrate it with a Fire-based\nmodule and a new Tiny module that includes max-pooling in each module. By\nstacking them we build networks that achieve the same accuracy as bigger ones,\nwhile reducing the number of parameters and considerably increasing\ncomputational performance. Our networks can classify a 96x96 sonar image with\n98.8 - 99.7 accuracy on only 41 to 61 milliseconds on a Raspberry Pi 2, which\ncorresponds to speedups of 28.6 - 19.7.\n",
"title": "Real-time convolutional networks for sonar image classification in low-power embedded systems"
}
| null | null | null | null | true | null |
9319
| null |
Default
| null | null |
null |
{
"abstract": " It is a crucial problem in robotics field to cage an object using robots like\nmultifingered hand. However the problem what is the caging for general\ngeometrical objects and robots has not been well-described in mathematics\nthough there were many rigorous studies on the methods how to cage an object by\ncertain robots. In this article, we investigate the caging problem more\nmathematically and describe the problem in terms of recursion of the simple\neuclidean moves. Using the description, we show that the caging has the degree\nof difficulty which is closely related to a combinatorial problem and a wire\npuzzle. It implies that in order to capture an object by caging, from a\npractical viewpoint the difficulty plays an important role.\n",
"title": "Mathematics in Caging of Robotics"
}
| null | null | null | null | true | null |
9320
| null |
Default
| null | null |
null |
{
"abstract": " Graph models are relevant in many fields, such as distributed computing,\nintelligent tutoring systems or social network analysis. In many cases, such\nmodels need to take changes in the graph structure into account, i.e. a varying\nnumber of nodes or edges. Predicting such changes within graphs can be expected\nto yield important insight with respect to the underlying dynamics, e.g. with\nrespect to user behaviour. However, predictive techniques in the past have\nalmost exclusively focused on single edges or nodes. In this contribution, we\nattempt to predict the future state of a graph as a whole. We propose to phrase\ntime series prediction as a regression problem and apply dissimilarity- or\nkernel-based regression techniques, such as 1-nearest neighbor, kernel\nregression and Gaussian process regression, which can be applied to graphs via\ngraph kernels. The output of the regression is a point embedded in a\npseudo-Euclidean space, which can be analyzed using subsequent dissimilarity-\nor kernel-based processing methods. We discuss strategies to speed up Gaussian\nProcesses regression from cubic to linear time and evaluate our approach on two\nwell-established theoretical models of graph evolution as well as two real data\nsets from the domain of intelligent tutoring systems. We find that simple\nregression methods, such as kernel regression, are sufficient to capture the\ndynamics in the theoretical models, but that Gaussian process regression\nsignificantly improves the prediction error for real-world data.\n",
"title": "Time Series Prediction for Graphs in Kernel and Dissimilarity Spaces"
}
| null | null | null | null | true | null |
9321
| null |
Default
| null | null |
null |
{
"abstract": " The metric space of phylogenetic trees defined by Billera, Holmes, and\nVogtmann, which we refer to as BHV space, provides a natural geometric setting\nfor describing collections of trees on the same set of taxa. However, it is\nsometimes necessary to analyze collections of trees on non-identical taxa sets\n(i.e., with different numbers of leaves), and in this context it is not evident\nhow to apply BHV space. Davidson et al. recently approached this problem by\ndescribing a combinatorial algorithm extending tree topologies to regions in\nhigher dimensional tree spaces, so that one can quickly compute which\ntopologies contain a given tree as partial data. In this paper, we refine and\nadapt their algorithm to work for metric trees to give a full characterization\nof the subspace of extensions of a subtree. We describe how to apply our\nalgorithm to define and search a space of possible supertrees and, for a\ncollection of tree fragments with different leaf sets, to measure their\ncompatibility.\n",
"title": "Geometric comparison of phylogenetic trees with different leaf sets"
}
| null | null | null | null | true | null |
9322
| null |
Default
| null | null |
null |
{
"abstract": " While social media offer great communication opportunities, they also\nincrease the vulnerability of young people to threatening situations online.\nRecent studies report that cyberbullying constitutes a growing problem among\nyoungsters. Successful prevention depends on the adequate detection of\npotentially harmful messages and the information overload on the Web requires\nintelligent systems to identify potential risks automatically. The focus of\nthis paper is on automatic cyberbullying detection in social media text by\nmodelling posts written by bullies, victims, and bystanders of online bullying.\nWe describe the collection and fine-grained annotation of a training corpus for\nEnglish and Dutch and perform a series of binary classification experiments to\ndetermine the feasibility of automatic cyberbullying detection. We make use of\nlinear support vector machines exploiting a rich feature set and investigate\nwhich information sources contribute the most for this particular task.\nExperiments on a holdout test set reveal promising results for the detection of\ncyberbullying-related posts. After optimisation of the hyperparameters, the\nclassifier yields an F1-score of 64% and 61% for English and Dutch\nrespectively, and considerably outperforms baseline systems based on keywords\nand word unigrams.\n",
"title": "Automatic Detection of Cyberbullying in Social Media Text"
}
| null | null | null | null | true | null |
9323
| null |
Default
| null | null |
null |
{
"abstract": " We discuss a cyclic cosmology in which the visible universe, or introverse,\nis all that is accessible to an observer while the extroverse represents the\ntotal spacetime originating from the time when the dark energy began to\ndominate. It is argued that entanglement entropy of the introverse is the more\nappropriate quantity to render infinitely cyclic, rather than the entropy of\nthe total universe. Since vanishing entanglement entropy implies disconnected\nspacetimes, at the turnaround when the introverse entropy is zero the\ndisconnected extroverse can be jettisoned with impunity.\n",
"title": "Holographic Entanglement Entropy in Cyclic Cosmology"
}
| null | null | null | null | true | null |
9324
| null |
Default
| null | null |
null |
{
"abstract": " Electronic health records (EHRs) have contributed to the computerization of\npatient records and can thus be used not only for efficient and systematic\nmedical services, but also for research on biomedical data science. However,\nthere are many missing values in EHRs when provided in matrix form, which is an\nimportant issue in many biomedical EHR applications. In this paper, we propose\na two-stage framework that includes missing data imputation and disease\nprediction to address the missing data problem in EHRs. We compared the disease\nprediction performance of generative adversarial networks (GANs) and\nconventional learning algorithms in combination with missing data prediction\nmethods. As a result, we obtained a level of accuracy of 0.9777, sensitivity of\n0.9521, specificity of 0.9925, area under the receiver operating characteristic\ncurve (AUC-ROC) of 0.9889, and F-score of 0.9688 with a stacked autoencoder as\nthe missing data prediction method and an auxiliary classifier GAN (AC-GAN) as\nthe disease prediction method. The comparison results show that a combination\nof a stacked autoencoder and an AC-GAN significantly outperforms other existing\napproaches. Our results suggest that the proposed framework is more robust for\ndisease prediction from EHRs with missing data.\n",
"title": "Adversarial Training for Disease Prediction from Electronic Health Records with Missing Data"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9325
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate the stability of a statistically stationary conductive state\nfor Rayleigh-Bénard convection between stress-free plates that arises due to\na bulk stochastic internal heating. This setup may be seen as a generalization\nto a stochastic setting of the seminal 1916 study of Lord Rayleigh. Our results\nindicate that stochastic forcing at small magnitude has a stabilizing effect,\nwhile strong stochastic forcing has a destabilizing effect. The methodology put\nforth in this article, which combines rigorous analysis with careful\ncomputation, also provides an approach to hydrodynamic stability for a variety\nof systems subject to a large scale stochastic forcing.\n",
"title": "Hydrodynamic stability in the presence of a stochastic forcing:a case study in convection"
}
| null | null |
[
"Physics",
"Mathematics"
] | null | true | null |
9326
| null |
Validated
| null | null |
null |
{
"abstract": " We construct a statistical indicator for the detection of short-term asset\nprice bubbles based on the information content of bid and ask market quotes for\nplain vanilla put and call options. Our construction makes use of the\nmartingale theory of asset price bubbles and the fact that such scenarios where\nthe price for an asset exceeds its fundamental value can in principle be\ndetected by analysis of the asymptotic behavior of the implied volatility\nsurface. For extrapolating this implied volatility, we choose the SABR model,\nmainly because of its decent fit to real option market quotes for a broad range\nof maturities and its ease of calibration. As main theoretical result, we show\nthat under lognormal SABR dynamics, we can compute a simple yet powerful\nclosed-form martingale defect indicator by solving an ill-posed inverse\ncalibration problem. In order to cope with the ill-posedness and to quantify\nthe uncertainty which is inherent to such an indicator, we adopt a Bayesian\nstatistical parameter estimation perspective. We probe the resulting posterior\ndensities with a combination of optimization and adaptive Markov chain Monte\nCarlo methods, thus providing a full-blown uncertainty estimation of all the\nunderlying parameters and the martingale defect indicator. Finally, we provide\nreal-market tests of the proposed option-based indicator with focus on tech\nstocks due to increasing concerns about a tech bubble 2.0.\n",
"title": "Asset Price Bubbles: An Option-based Indicator"
}
| null | null | null | null | true | null |
9327
| null |
Default
| null | null |
null |
{
"abstract": " One of the central notions to emerge from the study of persistent homology is\nthat of interleaving distance. It has found recent applications in symplectic\nand contact geometry, sheaf theory, computational geometry, and phylogenetics.\nHere we present a general study of this topic. We define interleaving of\nfunctors with common codomain as solutions to an extension problem. In order to\ndefine interleaving distance in this setting we are led to categorical\ngeneralizations of Hausdorff distance, Gromov-Hausdorff distance, and the space\nof metric spaces. We obtain comparisons with previous notions of interleaving\nvia the study of future equivalences. As an application we recover a definition\nof shift equivalences of discrete dynamical systems.\n",
"title": "Interleaving and Gromov-Hausdorff distance"
}
| null | null | null | null | true | null |
9328
| null |
Default
| null | null |
null |
{
"abstract": " The article investigates an evidence-based semantics for epistemic logics in\nwhich pieces of evidence are interpreted as equivalence relations on the\nepistemic worlds. It is shown that the properties of knowledge obtained from\npotentially infinitely many pieces of evidence are described by modal logic S5.\nAt the same time, the properties of knowledge obtained from only a finite\nnumber of pieces of evidence are described by modal logic S4. The main\ntechnical result is a sound and complete bi-modal logical system that describes\nproperties of these two modalities and their interplay.\n",
"title": "Attainable Knowledge"
}
| null | null | null | null | true | null |
9329
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the linear complementarity problems on extended\nsecond order cones. We convert a linear complementarity problem on an extended\nsecond order cone into a mixed complementarity problem on the non-negative\northant. We state necessary and sufficient conditions for a point to be a\nsolution of the converted problem. We also present solution strategies for this\nproblem, such as the Newton method and Levenberg-Marquardt algorithm. Finally,\nwe present some numerical examples.\n",
"title": "Linear complementarity problems on extended second order cones"
}
| null | null | null | null | true | null |
9330
| null |
Default
| null | null |
null |
{
"abstract": " We prove a Bernstein-von Mises theorem for a general class of high\ndimensional nonlinear Bayesian inverse problems in the vanishing noise limit.\nWe propose a sufficient condition on the growth rate of the number of unknown\nparameters under which the posterior distribution is asymptotically normal.\nThis growth condition is expressed explicitly in terms of the model dimension,\nthe degree of ill-posedness of the inverse problem and the noise parameter. The\ntheoretical results are applied to a Bayesian estimation of the medium\nparameter in an elliptic problem.\n",
"title": "On the Bernstein-Von Mises Theorem for High Dimensional Nonlinear Bayesian Inverse Problems"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
9331
| null |
Validated
| null | null |
null |
{
"abstract": " Word evolution refers to the changing meanings and associations of words\nthroughout time, as a byproduct of human language evolution. By studying word\nevolution, we can infer social trends and language constructs over different\nperiods of human history. However, traditional techniques such as word\nrepresentation learning do not adequately capture the evolving language\nstructure and vocabulary. In this paper, we develop a dynamic statistical model\nto learn time-aware word vector representation. We propose a model that\nsimultaneously learns time-aware embeddings and solves the resulting \"alignment\nproblem\". This model is trained on a crawled NYTimes dataset. Additionally, we\ndevelop multiple intuitive evaluation strategies of temporal word embeddings.\nOur qualitative and quantitative tests indicate that our method not only\nreliably captures this evolution over time, but also consistently outperforms\nstate-of-the-art temporal embedding approaches on both semantic accuracy and\nalignment quality.\n",
"title": "Dynamic Word Embeddings for Evolving Semantic Discovery"
}
| null | null | null | null | true | null |
9332
| null |
Default
| null | null |
null |
{
"abstract": " Fraenkel and Simpson showed that the number of distinct squares in a word of\nlength n is bounded from above by 2n, since at most two distinct squares have\ntheir rightmost, or last, occurrence begin at each position. Improvements by\nIlie to $2n-\\Theta(\\log n)$ and by Deza et al. to 11n/6 rely on the study of\ncombinatorics of FS-double-squares, when the maximum number of two last\noccurrences of squares begin. In this paper, we first study how to maximize\nruns of FS-double-squares in the prefix of a word. We show that for a given\npositive integer m, the minimum length of a word beginning with m\nFS-double-squares, whose lengths are equal, is 7m+3. We construct such a word\nand analyze its distinct-square-sequence as well as its\ndistinct-square-density. We then generalize our construction. We also construct\nwords with high distinct-square-densities that approach 5/6.\n",
"title": "Constructing Words with High Distinct Square Densities"
}
| null | null | null | null | true | null |
9333
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we propound a question on the annihilator of Koszul\nhomologies of a system of parameters of an almost complete intersection $R$.\nThe question can be stated in terms of the acyclicity of certain (finite)\nresidual approximation complexes whose $0$-th homologies are the residue field\nof $R$. We show that our question has an affirmative answer for certain almost\ncomplete intersection rings with small multiplicities, as well as for the\n$1$-th Koszul homology of any almost complete intersection. The statement about\nthe $1$-th Koszul homology is shown to be equivalent to the Monomial Conjecture\nand thus follows from its validity.\n",
"title": "Annihilators of Koszul Homologies and Almost Complete Intersections"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9334
| null |
Validated
| null | null |
null |
{
"abstract": " We analyze how the knowledge to autonomously handle one type of intersection,\nrepresented as a Deep Q-Network, translates to other types of intersections\n(tasks). We view intersection handling as a deep reinforcement learning\nproblem, which approximates the state action Q function as a deep neural\nnetwork. Using a traffic simulator, we show that directly copying a network\ntrained for one type of intersection to another type of intersection decreases\nthe success rate. We also show that when a network that is pre-trained on Task\nA and then is fine-tuned on a Task B, the resulting network not only performs\nbetter on the Task B than an network exclusively trained on Task A, but also\nretained knowledge on the Task A. Finally, we examine a lifelong learning\nsetting, where we train a single network on five different types of\nintersections sequentially and show that the resulting network exhibited\ncatastrophic forgetting of knowledge on previous tasks. This result suggests a\nneed for a long-term memory component to preserve knowledge.\n",
"title": "Analyzing Knowledge Transfer in Deep Q-Networks for Autonomously Handling Multiple Intersections"
}
| null | null | null | null | true | null |
9335
| null |
Default
| null | null |
null |
{
"abstract": " In this note we describe how some objects from generalized geometry appear in\nthe qualitative analysis and numerical simulation of mechanical systems. In\nparticular we discuss double vector bundles and Dirac structures. It turns out\nthat those objects can be naturally associated to systems with constraints --\nwe recall the mathematical construction in the context of so called implicit\nLagrangian systems. We explain how they can be used to produce new numerical\nmethods, that we call Dirac integrators.\nOn a test example of a simple pendulum in a gravity field we compare the\nDirac integrators with classical explicit and implicit methods, we pay special\nattention to conservation of constrains. Then, on a more advanced example of\nthe Ziegler column we show that the choice of numerical methods can indeed\naffect the conclusions of qualitative analysis of the dynamics of mechanical\nsystems. We also tell why we think that Dirac integrators are appropriate for\nthis kind of systems by explaining the relation with the notions of geometric\ndegree of non-conservativity and kinematic structural stability.\n",
"title": "From modelling of systems with constraints to generalized geometry and back to numerics"
}
| null | null | null | null | true | null |
9336
| null |
Default
| null | null |
null |
{
"abstract": " Recently experience replay is widely used in various deep reinforcement\nlearning (RL) algorithms, in this paper we rethink the utility of experience\nreplay. It introduces a new hyper-parameter, the memory buffer size, which\nneeds carefully tuning. However unfortunately the importance of this new\nhyper-parameter has been underestimated in the community for a long time. In\nthis paper we did a systematic empirical study of experience replay under\nvarious function representations. We showcase that a large replay buffer can\nsignificantly hurt the performance. Moreover, we propose a simple O(1) method\nto remedy the negative influence of a large replay buffer. We showcase its\nutility in both simple grid world and challenging domains like Atari games.\n",
"title": "A Deeper Look at Experience Replay"
}
| null | null | null | null | true | null |
9337
| null |
Default
| null | null |
null |
{
"abstract": " We study the behavior of the spectrum of the Dirac operator together with a\nsymmetric $W^{1, \\infty}$-potential on spin manifolds under a collapse of\ncodimension one with bounded sectional curvature and diameter. If there is an\ninduced spin structure on the limit space $N$ then there are convergent\neigenvalues which converge to the spectrum of a first order differential\noperator $D$ on $N$ together with a symmetric $W^{1,\\infty}$-potential. If $N$\nis orientable and the dimension of the limit space is even then $D$ is the\nDirac operator $D^N$ on $N$ and if the dimension of the limit space is odd,\nthen $D = D^N \\oplus -D^N$.\n",
"title": "Dirac operators with $W^{1,\\infty}$-potential under codimension one collapse"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9338
| null |
Validated
| null | null |
null |
{
"abstract": " We have studied the impact of low-frequency magnetic flux noise upon\nsuperconducting transmon qubits with various levels of tunability. We find that\nqubits with weaker tunability exhibit dephasing that is less sensitive to flux\nnoise. This insight was used to fabricate qubits where dephasing due to flux\nnoise was suppressed below other dephasing sources, leading to flux-independent\ndephasing times T2* ~ 15 us over a tunable range of ~340 MHz. Such tunable\nqubits have the potential to create high-fidelity, fault-tolerant qubit gates\nand fundamentally improve scalability for a quantum processor.\n",
"title": "Tunable Superconducting Qubits with Flux-Independent Coherence"
}
| null | null | null | null | true | null |
9339
| null |
Default
| null | null |
null |
{
"abstract": " We start with a Riemann-Hilbert problem (RHP) related to a BD.I-type\nsymmetric spaces $SO(2r+1)/S(O(2r-2s +1)\\otimes O(2s))$, $s\\geq 1$. We consider\ntwo Riemann-Hilbert problems: the first formulated on the real axis\n$\\mathbb{R}$ in the complex $\\lambda$-plane; the second one is formulated on\n$\\mathbb{R} \\oplus i\\mathbb{R}$. The first RHP for $s=1$ allows one to solve\nthe Kulish-Sklyanin (KS) model; the second RHP is relevant for a new type of KS\nmodel. An important example for nontrivial deep reductions of KS model is\ngiven. Its effect on the scattering matrix is formulated. In particular we\nobtain new 2-component NLS equations. Finally, using the Wronskian relations we\ndemonstrate that the inverse scattering method for KS models may be understood\nas a generalized Fourier transforms. Thus we have a tool to derive all their\nfundamental properties, including the hierarchy of equations and the hierarchy\nof their Hamiltonian structures.\n",
"title": "Kulish-Sklyanin type models: integrability and reductions"
}
| null | null | null | null | true | null |
9340
| null |
Default
| null | null |
null |
{
"abstract": " Extending results of Rais-Tauvel, Macedo-Savage, and Arakawa-Premet, we prove\nthat under mild restrictions on the Lie algebra $\\mathfrak q$ having the\npolynomial ring of symmetric invariants, the m-th Takiff algebra of $\\mathfrak\nq$, $\\mathfrak q\\langle m\\rangle$, also has a polynomial ring of symmetric\ninvariants.\n",
"title": "Takiff algebras with polynomial rings of symmetric invariants"
}
| null | null | null | null | true | null |
9341
| null |
Default
| null | null |
null |
{
"abstract": " Timing channels are a significant and growing security threat in computer\nsystems, with no established solution. We have recently argued that the OS must\nprovide time protection, in analogy to the established memory protection, to\nprotect applications from information leakage through timing channels. Based on\na recently-proposed implementation of time protection in the seL4 microkernel,\nwe investigate how such an implementation could be formally proved to prevent\ntiming channels. We postulate that this should be possible by reasoning about a\nhighly abstracted representation of the shared hardware resources that cause\ntiming channels.\n",
"title": "Can We Prove Time Protection?"
}
| null | null | null | null | true | null |
9342
| null |
Default
| null | null |
null |
{
"abstract": " We show that deciding whether a given graph $G$ of size $m$ has a unique\nperfect matching as well as finding that matching, if it exists, can be done in\ntime $O(m)$ if $G$ is either a cograph, or a split graph, or an interval graph,\nor claw-free. Furthermore, we provide a constructive characterization of the\nclaw-free graphs with a unique perfect matching.\n",
"title": "On some Graphs with a Unique Perfect Matching"
}
| null | null | null | null | true | null |
9343
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the optimal output-feedback control of a linear\ntime-invariant system where a stochastic event-based scheduler triggers the\ncommunication between the sensor and the controller. The primary goal of the\nuse of this type of scheduling strategy is to provide significant reductions in\nthe usage of the sensor-to-controller communication and, in turn, improve\nenergy expenditure in the network. In this paper, we aim to design an\nadmissible control policy, which is a function of the observed output, to\nminimize a quadratic cost function while employing a stochastic event-triggered\nscheduler that preserves the Gaussian property of the plant state and the\nestimation error. For the infinite horizon case, we present analytical\nexpressions that quantify the trade-off between the communication cost and\ncontrol performance of such event-triggered control systems. This trade-off is\nconfirmed quantitatively via numerical examples.\n",
"title": "Trade-Offs in Stochastic Event-Triggered Control"
}
| null | null | null | null | true | null |
9344
| null |
Default
| null | null |
null |
{
"abstract": " Graphene as a zero-bandgap two-dimensional semiconductor with a linear\nelectron band dispersion near the Dirac points has the potential to exhibit\nvery interesting nonlinear optical properties. In particular, third harmonic\ngeneration of terahertz radiation should occur due to the nonlinear\nrelationship between the crystal momentum and the current density. In this\nwork, we investigate the terahertz nonlinear response of graphene inside a\nparallel-plate waveguide. We optimize the plate separation and Fermi energy of\nthe graphene to maximize third harmonic generation, by maximizing the nonlinear\ninteraction while minimizing the loss and phase mismatch. The results obtained\nshow an increase by more than a factor of 100 in the power efficiency relative\nto a normal-incidence configuration for a 2 terahertz incident field.\n",
"title": "Third Harmonic THz Generation from Graphene in a Parallel-Plate Waveguide"
}
| null | null |
[
"Physics"
] | null | true | null |
9345
| null |
Validated
| null | null |
null |
{
"abstract": " Path integrals describing quantum many-body systems can be calculated with\nMonte Carlo sampling techniques, but average quantities are often subject to\nsignal-to-noise ratios that degrade exponentially with time. A\nphase-reweighting technique inspired by recent observations of random walk\nstatistics in correlation functions is proposed that allows energy levels to be\nextracted from late-time correlation functions with time-independent\nsignal-to-noise ratios. Phase reweighting effectively includes dynamical\nrefinement of source magnitudes but introduces a bias associated with the\nphase. This bias can be removed by performing an extrapolation, but at the\nexpense of re-introducing a signal-to-noise problem. Lattice Quantum\nChromodynamics calculations of the $\\rho$ and nucleon masses and of the\n$\\Xi\\Xi$ binding energy show consistency between standard results obtained\nusing earlier-time correlation functions and phase-reweighted results using\nlate-time correlation functions inaccessible to standard statistical analysis\nmethods.\n",
"title": "Taming the Signal-to-Noise Problem in Lattice QCD by Phase Reweighting"
}
| null | null | null | null | true | null |
9346
| null |
Default
| null | null |
null |
{
"abstract": " The BCML system is a beam monitoring device in the CMS experiment at the LHC.\nAs detectors poly-crystalline diamond sensors are used. Here high particle\nrates occur from the colliding beams scattering particles outside the beam\npipe. These particles cause defects, which act as traps for the ionization,\nthus reducing the CCE. However, the loss in CCE was much more severe than\nexpected. The reason why in real experiments the CCE is so much worse than in\nlaboratory experiments is related to the rate of incident particles. At high\nparticle rates the trapping rate of the ionization is so high compared with the\ndetrapping rate, that space charge builds up. This space charge reduces locally\nthe internal electric field, which in turn increases the trapping rate and\nhence reduces the CCE even further. In order to connect these macroscopic\nmeasurements with the microscopic defects acting as traps for the ionization\ncharge the TCAD simulation program SILVACO was used. Two effective acceptor and\ndonor levels were needed to fit the data. Using this effective defect model the\nhighly non- linear rate dependent diamond polarization as function of the\nparticle rate environment and the resulting signal loss could be simulated.\n",
"title": "Description of radiation damage in diamond sensors using an effective defect model"
}
| null | null | null | null | true | null |
9347
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we deal with time-invariant spatially coupled low-density\nparity-check convolutional codes (SC-LDPC-CCs). Classic design approaches\nusually start from quasi-cyclic low-density parity-check (QC-LDPC) block codes\nand exploit suitable unwrapping procedures to obtain SC-LDPC-CCs. We show that\nthe direct design of the SC-LDPC-CCs syndrome former matrix or, equivalently,\nthe symbolic parity-check matrix, leads to codes with smaller syndrome former\nconstraint lengths with respect to the best solutions available in the\nliterature. We provide theoretical lower bounds on the syndrome former\nconstraint length for the most relevant families of SC-LDPC-CCs, under\nconstraints on the minimum length of cycles in their Tanner graphs. We also\npropose new code design techniques that approach or achieve such theoretical\nlimits.\n",
"title": "Design and Analysis of Time-Invariant SC-LDPC Convolutional Codes With Small Constraint Length"
}
| null | null | null | null | true | null |
9348
| null |
Default
| null | null |
null |
{
"abstract": " The prediction of organic reaction outcomes is a fundamental problem in\ncomputational chemistry. Since a reaction may involve hundreds of atoms, fully\nexploring the space of possible transformations is intractable. The current\nsolution utilizes reaction templates to limit the space, but it suffers from\ncoverage and efficiency issues. In this paper, we propose a template-free\napproach to efficiently explore the space of product molecules by first\npinpointing the reaction center -- the set of nodes and edges where graph edits\noccur. Since only a small number of atoms contribute to reaction center, we can\ndirectly enumerate candidate products. The generated candidates are scored by a\nWeisfeiler-Lehman Difference Network that models high-order interactions\nbetween changes occurring at nodes across the molecule. Our framework\noutperforms the top-performing template-based approach with a 10\\% margin,\nwhile running orders of magnitude faster. Finally, we demonstrate that the\nmodel accuracy rivals the performance of domain experts.\n",
"title": "Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
9349
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, Legendre curves on unit tangent bundle are given using\nrotation minimizing (RM) vector fields. Ruled surfaces corresponding to these\ncurves are represented. Singularities of these ruled surfaces are also analyzed\nand classifed.\n",
"title": "Legendre curves and singularities of a ruled surface according to rotation minimizing frame"
}
| null | null | null | null | true | null |
9350
| null |
Default
| null | null |
null |
{
"abstract": " Giant vortices with higher phase-winding than $2\\pi$ are usually\nenergetically unfavorable, but geometric symmetry constraints on a\nsuperconductor in a magnetic field are known to stabilize such objects. Here,\nwe show via microscopic calculations that giant vortices can appear in\nintrinsically non-superconducting materials, even without any applied magnetic\nfield. The enabling mechanism is the proximity effect to a host superconductor\nwhere a current flows, and we also demonstrate that antivortices can appear in\nthis setup. Our results open the possibility to study electrically controllable\ntopological defects in unusual environments, which do not have to be exposed to\nmagnetic fields or intrinsically superconducting, but instead display other\ntypes of order.\n",
"title": "Field-free nucleation of antivortices and giant vortices in non-superconducting materials"
}
| null | null | null | null | true | null |
9351
| null |
Default
| null | null |
null |
{
"abstract": " We study vortex patterns in a prototype nonlinear optical system:\ncounterpropagating laser beams in a photorefractive crystal, with or without\nthe background photonic lattice. The vortices are effectively planar and\ndescribed by the winding number and the \"flavor\" index, stemming from the fact\nthat we have two parallel beams propagating in opposite directions. The problem\nis amenable to the methods of statistical field theory and generalizes the\nBerezinsky-Kosterlitz-Thouless transition of the XY model to the \"two-flavor\"\ncase. In addition to the familiar conductor and insulator phases, we also have\nthe perfect conductor (vortex proliferation in both beams/\"flavors\") and the\nfrustrated insulator (energy costs of vortex proliferation and vortex\nannihilation balance each other). In the presence of disorder in the background\nlattice, a novel phase appears which shows long-range correlations and absence\nof long-range order, thus being analogous to spin glasses. An important benefit\nof this approach is that qualitative behavior of patterns can be known without\nintensive numerical work over large areas of the parameter space. More\ngenerally, we would like to draw attention to connections between the\n(classical) pattern-forming systems in photorefractive optics and the methods\nof (quantum) condensed matter and field theory: on one hand, we use the\nfield-theoretical methods (renormalization group, replica formalism) to analyze\nthe patterns; on the other hand, the observed phases are analogous to those\nseen in magnetic systems, and make photorefractive optics a fruitful testing\nground for condensed matter systems. As an example, we map our system to a\ndoped $O(3)$ antiferromagnet with $\\mathbb{Z}_2$ defects, which has the same\nstructure of the phase diagram.\n",
"title": "Quantum criticality in photorefractive optics: vortices in laser beams and antiferromagnets"
}
| null | null | null | null | true | null |
9352
| null |
Default
| null | null |
null |
{
"abstract": " Recent work on imitation learning has generated policies that reproduce\nexpert behavior from multi-modal data. However, past approaches have focused\nonly on recreating a small number of distinct, expert maneuvers, or have relied\non supervised learning techniques that produce unstable policies. This work\nextends InfoGAIL, an algorithm for multi-modal imitation learning, to reproduce\nbehavior over an extended period of time. Our approach involves reformulating\nthe typical imitation learning setting to include \"burn-in demonstrations\" upon\nwhich policies are conditioned at test time. We demonstrate that our approach\noutperforms standard InfoGAIL in maximizing the mutual information between\npredicted and unseen style labels in road scene simulations, and we show that\nour method leads to policies that imitate expert autonomous driving systems\nover long time horizons.\n",
"title": "Burn-In Demonstrations for Multi-Modal Imitation Learning"
}
| null | null | null | null | true | null |
9353
| null |
Default
| null | null |
null |
{
"abstract": " We introduce seven families of stochastic systems of interacting particles in\none-dimension corresponding to the seven families of irreducible reduced affine\nroot systems. We prove that they are determinantal in the sense that all\nspatio-temporal correlation functions are given by determinants controlled by a\nsingle function called the spatio-temporal correlation kernel. For the four\nfamilies ${A}_{N-1}$, ${B}_N$, ${C}_N$ and ${D}_N$, we identify the systems of\nstochastic differential equations solved by these determinantal processes,\nwhich will be regarded as the elliptic extensions of the Dyson model. Here we\nuse the notion of martingales in probability theory and the elliptic\ndeterminant evaluations of the Macdonald denominators of irreducible reduced\naffine root systems given by Rosengren and Schlosser.\n",
"title": "Elliptic Determinantal Processes and Elliptic Dyson Models"
}
| null | null | null | null | true | null |
9354
| null |
Default
| null | null |
null |
{
"abstract": " Recent studies have shown that close-in brown dwarfs in the mass range 35-55\nM$_{\\rm Jup}$ are almost depleted as companions to stars, suggesting that\nobjects with masses above and below this gap might have different formation\nmechanisms. We determine the fundamental stellar parameters, as well as\nindividual abundances for a large sample of stars known to have a substellar\ncompanion in the brown dwarf regime. The sample is divided into stars hosting\n\"massive\" and \"low-mass\" brown dwarfs. Following previous works a threshold of\n42.5 M$_{\\rm Jup}$ was considered. Our results confirm that stars with brown\ndwarf companions do not follow the well-established gas-giant planet\nmetallicity correlation seen in main-sequence planet hosts. Stars harbouring\n\"massive\" brown dwarfs show similar metallicity and abundance distribution as\nstars without known planets or with low-mass planets. We find a tendency of\nstars harbouring \"less-massive\" brown dwarfs of having slightly larger\nmetallicity, [X$_{\\rm Fe}$/Fe] values, and abundances of Sc II, Mn I, and Ni I\nin comparison with the stars having the massive brown dwarfs. The data suggest,\nas previously reported, that massive and low-mass brown dwarfs might present\ndifferences in period and eccentricity. We find evidence of a non-metallicity\ndependent mechanism for the formation of massive brown dwarfs. Our results\nagree with a scenario in which massive brown dwarfs are formed as stars. At\nhigh-metallicities, the core-accretion mechanism might become efficient in the\nformation of low-mass brown dwarfs while at lower metallicities low-mass brown\ndwarfs could form by gravitational instability in turbulent protostellar discs.\n",
"title": "Searching for chemical signatures of brown dwarf formation"
}
| null | null | null | null | true | null |
9355
| null |
Default
| null | null |
null |
{
"abstract": " The Dirac equation requires a treatment of the step potential that differs\nfundamentally from the traditional treatment, because the Dirac plane waves,\nbesides momentum and spin, are characterized by a quantum number with the\nphysical meaning of sign of charge. Since the Hermitean operator corresponding\nto this quantum number does not commute with the step potential, the time\ndisplacement parameter used in the ansatz of the stationary state does not have\nthe physical meaning of energy. Therefore there are no paradoxal values of the\nenergy. The new solution of the Dirac equation with a step potential is\nobtained. This solution, again, allows for phenomena of the Klein paradox type,\nbut in addition it contains a positron amplitude localized at the threshold\npoint of the step potential.\n",
"title": "The Klein Paradox: A New Treatment"
}
| null | null | null | null | true | null |
9356
| null |
Default
| null | null |
null |
{
"abstract": " In recent years, real estate industry has captured government and public\nattention around the world. The factors influencing the prices of real estate\nare diversified and complex. However, due to the limitations and one-sidedness\nof their respective views, they did not provide enough theoretical basis for\nthe fluctuation of house price and its influential factors. The purpose of this\npaper is to build a housing price model to make the scientific and objective\nanalysis of London's real estate market trends from the year 1996 to 2016 and\nproposes some countermeasures to reasonably control house prices. Specifically,\nthe paper analyzes eight factors which affect the house prices from two\naspects: housing supply and demand and find out the factor which is of vital\nimportance to the increase of housing price per square meter. The problem of a\nhigh level of multicollinearity between them is solved by using principal\ncomponents analysis.\n",
"title": "What are the most important factors that influence the changes in London Real Estate Prices? How to quantify them?"
}
| null | null | null | null | true | null |
9357
| null |
Default
| null | null |
null |
{
"abstract": " We present new Atacama Large Millimeter/sub-millimeter Array (ALMA) 1.3 mm\ncontinuum observations of the SR 24S transition disk with an angular resolution\n$\\lesssim0.18\"$ (12 au radius). We perform a multi-wavelength investigation by\ncombining new data with previous ALMA data at 0.45 mm. The visibilities and\nimages of the continuum emission at the two wavelengths are well characterized\nby a ring-like emission. Visibility modeling finds that the ring-like emission\nis narrower at longer wavelengths, in good agreement with models of dust\ntrapping in pressure bumps, although there are complex residuals that suggest\npotentially asymmetric structures. The 0.45 mm emission has a shallower profile\ninside the central cavity than the 1.3 mm emission. In addition, we find that\nthe $^{13}$CO and C$^{18}$O (J=2-1) emission peaks at the center of the\ncontinuum cavity. We do not detect either continuum or gas emission from the\nnorthern companion to this system (SR 24N), which is itself a binary system.\nThe upper limit for the dust disk mass of SR 24N is $\\lesssim\n0.12\\,M_{\\bigoplus}$, which gives a disk mass ratio in dust between the two\ncomponents of $M_{\\mathrm{dust, SR\\,24S}}/M_{\\mathrm{dust,\nSR\\,24N}}\\gtrsim840$. The current ALMA observations may imply that either\nplanets have already formed in the SR 24N disk or that dust growth to mm-sizes\nis inhibited there and that only warm gas, as seen by ro-vibrational CO\nemission inside the truncation radii of the binary, is present.\n",
"title": "A Multi-Wavelength Analysis of Dust and Gas in the SR 24S Transition Disk"
}
| null | null |
[
"Physics"
] | null | true | null |
9358
| null |
Validated
| null | null |
null |
{
"abstract": " We report for the first time the observation of bunching of monoatomic steps\non vicinal W(110) surfaces induced by step up or step down currents across the\nsteps. Measurements reveal that the size scaling exponent {\\gamma}, connecting\nthe maximal slope of a bunch with its height, differs depending on the current\ndirection. We provide a numerical perspective by using an atomistic scale model\nwith a conserved surface flux to mimic experimental conditions, and also for\nthe first time show that there is an interval of parameters in which the\nvicinal surface is unstable against step bunching for both directions of the\nadatom drift.\n",
"title": "Step bunching with both directions of the current: Vicinal W(110) surfaces versus atomistic scale model"
}
| null | null | null | null | true | null |
9359
| null |
Default
| null | null |
null |
{
"abstract": " Bandit is a framework for designing sequential experiments. In each\nexperiment, a learner selects an arm $A \\in \\mathcal{A}$ and obtains an\nobservation corresponding to $A$. Theoretically, the tight regret lower-bound\nfor the general bandit is polynomial with respect to the number of arms\n$|\\mathcal{A}|$. This makes bandit incapable of handling an exponentially large\nnumber of arms, hence the bandit problem with side-information is often\nconsidered to overcome this lower bound. Recently, a bandit framework over a\ncausal graph was introduced, where the structure of the causal graph is\navailable as side-information. A causal graph is a fundamental model that is\nfrequently used with a variety of real problems. In this setting, the arms are\nidentified with interventions on a given causal graph, and the effect of an\nintervention propagates throughout all over the causal graph. The task is to\nfind the best intervention that maximizes the expected value on a target node.\nExisting algorithms for causal bandit overcame the\n$\\Omega(\\sqrt{|\\mathcal{A}|/T})$ simple-regret lower-bound; however, their\nalgorithms work only when the interventions $\\mathcal{A}$ are localized around\na single node (i.e., an intervention propagates only to its neighbors).\nWe propose a novel causal bandit algorithm for an arbitrary set of\ninterventions, which can propagate throughout the causal graph. We also show\nthat it achieves $O(\\sqrt{ \\gamma^*\\log(|\\mathcal{A}|T) / T})$ regret bound,\nwhere $\\gamma^*$ is determined by using a causal graph structure. In\nparticular, if the in-degree of the causal graph is bounded, then $\\gamma^* =\nO(N^2)$, where $N$ is the number $N$ of nodes.\n",
"title": "Causal Bandits with Propagating Inference"
}
| null | null |
[
"Statistics"
] | null | true | null |
9360
| null |
Validated
| null | null |
null |
{
"abstract": " We present a new algorithm for the 2D Sliding Window Discrete Fourier\nTransform (SWDFT). Our algorithm avoids repeating calculations in overlapping\nwindows by storing them in a tree data-structure based on the ideas of the\nCooley- Tukey Fast Fourier Transform (FFT). For an $N_0 \\times N_1$ array and\n$n_0 \\times n_1$ windows, our algorithm takes $O(N_0 N_1 n_0 n_1)$ operations.\nWe provide a C implementation of our algorithm for the Radix-2 case, compare\nours with existing algorithms, and show how our algorithm easily extends to\nhigher dimensions.\n",
"title": "The 2D Tree Sliding Window Discrete Fourier Transform"
}
| null | null | null | null | true | null |
9361
| null |
Default
| null | null |
null |
{
"abstract": " In this study, we investigate the limits of the current state of the art AI\nsystem for detecting buffer overflows and compare it with current static\nanalysis tools. To do so, we developed a code generator, s-bAbI, capable of\nproducing an arbitrarily large number of code samples of controlled complexity.\nWe found that the static analysis engines we examined have good precision, but\npoor recall on this dataset, except for a sound static analyzer that has good\nprecision and recall. We found that the state of the art AI system, a memory\nnetwork modeled after Choi et al. [1], can achieve similar performance to the\nstatic analysis engines, but requires an exhaustive amount of training data in\norder to do so. Our work points towards future approaches that may solve these\nproblems; namely, using representations of code that can capture appropriate\nscope information and using deep learning methods that are able to perform\narithmetic operations.\n",
"title": "Towards security defect prediction with AI"
}
| null | null |
[
"Statistics"
] | null | true | null |
9362
| null |
Validated
| null | null |
null |
{
"abstract": " In general, neural networks are not currently capable of learning tasks in a\nsequential fashion. When a novel, unrelated task is learnt by a neural network,\nit substantially forgets how to solve previously learnt tasks. One of the\noriginal solutions to this problem is pseudo-rehearsal, which involves learning\nthe new task while rehearsing generated items representative of the previous\ntask/s. This is very effective for simple tasks. However, pseudo-rehearsal has\nnot yet been successfully applied to very complex tasks because in these tasks\nit is difficult to generate representative items. We accomplish\npseudo-rehearsal by using a Generative Adversarial Network to generate items so\nthat our deep network can learn to sequentially classify the CIFAR-10, SVHN and\nMNIST datasets. After training on all tasks, our network loses only 1.67%\nabsolute accuracy on CIFAR-10 and gains 0.24% absolute accuracy on SVHN. Our\nmodel's performance is a substantial improvement compared to the current state\nof the art solution.\n",
"title": "Pseudo-Recursal: Solving the Catastrophic Forgetting Problem in Deep Neural Networks"
}
| null | null | null | null | true | null |
9363
| null |
Default
| null | null |
null |
{
"abstract": " While scale invariance is commonly observed in each component of real world\nmultivariate signals, it is also often the case that the inter-component\ncorrelation structure is not fractally connected, i.e., its scaling behavior is\nnot determined by that of the individual components. To model this situation in\na versatile manner, we introduce a class of multivariate Gaussian stochastic\nprocesses called Hadamard fractional Brownian motion (HfBm). Its theoretical\nstudy sheds light on the issues raised by the joint requirement of entry-wise\nscaling and departures from fractal connectivity. An asymptotically normal\nwavelet-based estimator for its scaling parameter, called the Hurst matrix, is\nproposed, as well as asymptotically valid confidence intervals. The latter are\naccompanied by original finite sample procedures for computing confidence\nintervals and testing fractal connectivity from one single and finite size\nobservation. Monte Carlo simulation studies are used to assess the estimation\nperformance as a function of the (finite) sample size, and to quantify the\nimpact of omitting wavelet cross-correlation terms. The simulation studies are\nshown to validate the use of approximate confidence intervals, together with\nthe significance level and power of the fractal connectivity test. The test\nperformance and properties are further studied as functions of the HfBm\nparameters.\n",
"title": "Multivariate Hadamard self-similarity: testing fractal connectivity"
}
| null | null | null | null | true | null |
9364
| null |
Default
| null | null |
null |
{
"abstract": " We present an easy-to-implement and efficient analytical inversion algorithm\nfor the unbiased random sampling of a set of points on a triangle mesh whose\nsurface density is specified by barycentric interpolation of non-negative\nper-vertex weights. The correctness of the inversion algorithm is verified via\nstatistical tests, and we show that it is faster on average than rejection\nsampling.\n",
"title": "Efficient barycentric point sampling on meshes"
}
| null | null | null | null | true | null |
9365
| null |
Default
| null | null |
null |
{
"abstract": " Superregular (SR) breathers are nonlinear wave structures formed by a unique\nnonlinear superposition of pairs of quasi-Akhmediev breathers. They describe a\ncomplete scenario of modulation instability that develops from localized small\nperturbations as well as an unusual quasiannihilation of breather collision.\nHere, we demonstrate that femtosecond optical SR breathers in optical fibers\nexhibit intriguing half-transition and full-suppression states, which are\nabsent in the picosecond regime governed by the standard nonlinear\nSchrödinger equation. In particular, the full-suppression mode, which is\nstrictly associated with the regime of vanishing growth rate of modulation\ninstability, reveals a crucial \\textit{non-amplifying} nonlinear dynamics of\nlocalized small perturbations. We numerically confirm the robustness of such\ndifferent SR modes excited from ideal and nonideal initial states in both\nintegrable and nonintegrable cases.\n",
"title": "Femtosecond Optical Superregular Breathers"
}
| null | null |
[
"Physics"
] | null | true | null |
9366
| null |
Validated
| null | null |
null |
{
"abstract": " Real network datasets provide significant benefits for understanding\nphenomena such as information diffusion or network evolution. Yet the privacy\nrisks raised from sharing real graph datasets, even when stripped of user\nidentity information, are significant. When nodes have associated attributes,\nthe privacy risks increase. In this paper we quantitatively study the impact of\nbinary node attributes on node privacy by employing machine-learning-based\nre-identification attacks and exploring the interplay between graph topology\nand attribute placement. Our experiments show that the population's diversity\non the binary attribute consistently degrades anonymity.\n",
"title": "Diversity, Topology, and the Risk of Node Re-identification in Labeled Social Graphs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9367
| null |
Validated
| null | null |
null |
{
"abstract": " We define two algebra automorphisms $T_0$ and $T_1$ of the $q$-Onsager\nalgebra $B_c$, which provide an analog of G. Lusztig's braid group action for\nquantum groups. These automorphisms are used to define root vectors which give\nrise to a PBW basis for $B_c$. We show that the root vectors satisfy\n$q$-analogs of Onsager's original commutation relations. The paper is much\ninspired by I. Damiani's construction and investigation of root vectors for the\nquantized enveloping algebra of $\\widehat{\\mathfrak{sl}}_2$.\n",
"title": "Braid group action and root vectors for the $q$-Onsager algebra"
}
| null | null | null | null | true | null |
9368
| null |
Default
| null | null |
null |
{
"abstract": " We examine Lagrangian techniques for computing underapproximations of\nfinite-time horizon, stochastic reach-avoid level-sets for discrete-time,\nnonlinear systems. We use the concept of reachability of a target tube in the\ncontrol literature to define robust reach-avoid sets which are parameterized by\nthe target set, safe set, and the set in which the disturbance is drawn from.\nWe unify two existing Lagrangian approaches to compute these sets and establish\nthat there exists an optimal control policy of the robust reach-avoid sets\nwhich is a Markov policy. Based on these results, we characterize the subset of\nthe disturbance space whose corresponding robust reach-avoid set for the given\ntarget and safe set is a guaranteed underapproximation of the stochastic\nreach-avoid level-set of interest. The proposed approach dramatically improves\nthe computational efficiency for obtaining an underapproximation of stochastic\nreach-avoid level-sets when compared to the traditional approaches based on\ngridding. Our method, while conservative, does not rely on a grid, implying\nscalability as permitted by the known computational geometry constraints. We\ndemonstrate the method on two examples: a simple two-dimensional integrator,\nand a space vehicle rendezvous-docking problem.\n",
"title": "Underapproximation of Reach-Avoid Sets for Discrete-Time Stochastic Systems via Lagrangian Methods"
}
| null | null | null | null | true | null |
9369
| null |
Default
| null | null |
null |
{
"abstract": " Domain-specific languages (DSLs) are of increasing importance in scientific\nhigh-performance computing to reduce development costs, raise the level of\nabstraction and, thus, ease scientific programming. However, designing and\nimplementing DSLs is not an easy task, as it requires knowledge of the\napplication domain and experience in language engineering and compilers.\nConsequently, many DSLs follow a weak approach using macros or text generators,\nwhich lack many of the features that make a DSL a comfortable for programmers.\nSome of these features---e.g., syntax highlighting, type inference, error\nreporting, and code completion---are easily provided by language workbenches,\nwhich combine language engineering techniques and tools in a common ecosystem.\nIn this paper, we present the Parallel Particle-Mesh Environment (PPME), a DSL\nand development environment for numerical simulations based on particle methods\nand hybrid particle-mesh methods. PPME uses the meta programming system (MPS),\na projectional language workbench. PPME is the successor of the Parallel\nParticle-Mesh Language (PPML), a Fortran-based DSL that used conventional\nimplementation strategies. We analyze and compare both languages and\ndemonstrate how the programmer's experience can be improved using static\nanalyses and projectional editing. Furthermore, we present an explicit domain\nmodel for particle abstractions and the first formal type system for particle\nmethods.\n",
"title": "A Domain-Specific Language and Editor for Parallel Particle Methods"
}
| null | null | null | null | true | null |
9370
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the Nystr{ö}m type subsampling for large scale\nkernel methods to reduce the computational complexities of big data. We discuss\nthe multi-penalty regularization scheme based on Nystr{ö}m type subsampling\nwhich is motivated from well-studied manifold regularization schemes. We\ndevelop a theoretical analysis of multi-penalty least-square regularization\nscheme under the general source condition in vector-valued function setting,\ntherefore the results can also be applied to multi-task learning problems. We\nachieve the optimal minimax convergence rates of multi-penalty regularization\nusing the concept of effective dimension for the appropriate subsampling size.\nWe discuss an aggregation approach based on linear function strategy to combine\nvarious Nystr{ö}m approximants. Finally, we demonstrate the performance of\nmulti-penalty regularization based on Nystr{ö}m type subsampling on\nCaltech-101 data set for multi-class image classification and NSL-KDD benchmark\ndata set for intrusion detection problem.\n",
"title": "Manifold regularization based on Nystr{ö}m type subsampling"
}
| null | null | null | null | true | null |
9371
| null |
Default
| null | null |
null |
{
"abstract": " In this work, a novel ring polymer representation for multi-level quantum\nsystem is proposed for thermal average calculations. The proposed presentation\nkeeps the discreteness of the electronic states: besides position and momentum,\neach bead in the ring polymer is also characterized by a surface index\nindicating the electronic energy surface. A path integral molecular dynamics\nwith surface hopping (PIMD-SH) dynamics is also developed to sample the\nequilibrium distribution of ring polymer configurational space. The PIMD-SH\nsampling method is validated theoretically and by numerical examples.\n",
"title": "Path integral molecular dynamics with surface hopping for thermal equilibrium sampling of nonadiabatic systems"
}
| null | null | null | null | true | null |
9372
| null |
Default
| null | null |
null |
{
"abstract": " The nonlinear thin-shell instability (NTSI) may explain some of the turbulent\nhydrodynamic structures that are observed close to the collision boundary of\nenergetic astrophysical outflows. It develops in nonplanar shells that are\nbounded on either side by a hydrodynamic shock, provided that the amplitude of\nthe seed oscillations is sufficiently large. The hydrodynamic NTSI has a\nmicroscopic counterpart in collisionless plasma. A sinusoidal displacement of a\nthin shell, which is formed by the collision of two clouds of unmagnetized\nelectrons and protons, grows and saturates on timescales of the order of the\ninverse proton plasma frequency. Here we increase the wavelength of the seed\nperturbation by a factor 4 compared to that in a previous study. Like in the\ncase of the hydrodynamic NTSI, the increase in the wavelength reduces the\ngrowth rate of the microscopic NTSI. The prolonged growth time of the\nmicroscopic NTSI allows the waves, which are driven by the competing ion\nacoustic instability, to grow to a large amplitude before the NTSI saturates\nand they disrupt the latter. The ion acoustic instability thus imposes a limit\non the largest wavelength that can be destabilized by the NTSI in collisionless\nplasma. The limit can be overcome by binary collisions. We bring forward\nevidence for an overstability of the collisionless NTSI.\n",
"title": "The interplay of the collisionless nonlinear thin-shell instability with the ion acoustic instability"
}
| null | null | null | null | true | null |
9373
| null |
Default
| null | null |
null |
{
"abstract": " The multivariate contaminated normal (MCN) distribution represents a simple\nheavy-tailed generalization of the multivariate normal (MN) distribution to\nmodel elliptical contoured scatters in the presence of mild outliers, referred\nto as \"bad\" points. The MCN can also automatically detect bad points. The price\nof these advantages is two additional parameters, both with specific and useful\ninterpretations: proportion of good observations and degree of contamination.\nHowever, points may be bad in some dimensions but good in others. The use of an\noverall proportion of good observations and of an overall degree of\ncontamination is limiting. To overcome this limitation, we propose a multiple\nscaled contaminated normal (MSCN) distribution with a proportion of good\nobservations and a degree of contamination for each dimension. Once the model\nis fitted, each observation has a posterior probability of being good with\nrespect to each dimension. Thanks to this probability, we have a method for\nsimultaneous directional robust estimation of the parameters of the MN\ndistribution based on down-weighting and for the automatic directional\ndetection of bad points by means of maximum a posteriori probabilities. The\nterm \"directional\" is added to specify that the method works separately for\neach dimension. Mixtures of MSCN distributions are also proposed as an\napplication of the proposed model for robust clustering. An extension of the EM\nalgorithm is used for parameter estimation based on the maximum likelihood\napproach. Real and simulated data are used to show the usefulness of our\nmixture with respect to well-established mixtures of symmetric distributions\nwith heavy tails.\n",
"title": "Multiple Scaled Contaminated Normal Distribution and Its Application in Clustering"
}
| null | null | null | null | true | null |
9374
| null |
Default
| null | null |
null |
{
"abstract": " We present new determinations of the stellar-to-halo mass relation (SHMR) at\n$z=0-10$ that match the evolution of the galaxy stellar mass function, the\nSFR$-M_*$ relation,and the cosmic star formation rate. We utilize a compilation\nof 40 observational studies from the literature and correct them for potential\nbiases. Using our robust determinations of halo mass assembly and the SHMR, we\ninfer star formation histories, merger rates, and structural properties for\naverage galaxies, combining star-forming and quenched galaxies. Our main\nfindings: (1) The halo mass $M_{50}$ above which 50\\% of galaxies are quenched\ncoincides with sSFR/sMAR$\\sim1$, where sMAR is the specific halo mass accretion\nrate. (2) $M_{50}$ increases with redshift, presumably due to cold streams\nbeing more efficient at high redshift while virial shocks and AGN feedback\nbecome more relevant at lower redshifts. (3) The ratio sSFR/sMAR has a peak\nvalue, which occurs around $M_{\\rm vir}\\sim2\\times10^{11}M_{\\odot}$. (4) The\nstellar mass density within 1 kpc, $\\Sigma_1$, is a good indicator of the\ngalactic global sSFR. (5) Galaxies are statistically quenched after they reach\na maximum in $\\Sigma_1$, consistent with theoretical expectations of the gas\ncompaction model; this maximum depends on redshift. (6) In-situ star formation\nis responsible for most galactic stellar mass growth, especially for lower-mass\ngalaxies. (7) Galaxies grow inside out. The marked change in the slope of the\nsize--mass relation when galaxies became quenched, from $d\\log R_{\\rm\neff}/d\\log M_*\\sim0.35$ to $\\sim2.5$, could be the result of dry minor mergers.\n",
"title": "The Galaxy-Halo Connection Over The Last 13.3 Gyrs"
}
| null | null | null | null | true | null |
9375
| null |
Default
| null | null |
null |
{
"abstract": " This article proposes in depth comparative study of the most popular, used\nand analyzed Trust and Reputation System (TRS) according to the trust and\nreputation literature and in terms of specific trustworthiness criteria. This\nsurvey is realized relying on a selection of trustworthiness criteria that\nanalyze and evaluate the maturity and effectiveness of TRS. These criteria\ndescribe the utility, the usability, the performance and the effectiveness of\nthe TRS. We also provide a summary table of the compared TRS within a detailed\nand granular selection of trust and reputation aspects.\n",
"title": "State of the art of Trust and Reputation Systems in E-Commerce Context"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9376
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the problem of packing a family of disks \"on a shelf\", that is,\nsuch that each disk touches the $x$-axis from above and such that no two disks\noverlap. We prove that the problem of minimizing the distance between the\nleftmost point and the rightmost point of any disk is NP-hard. On the positive\nside, we show how to approximate this problem within a factor of 4/3 in $O(n\n\\log n)$ time, and provide an $O(n \\log n)$-time exact algorithm for a special\ncase, in particular when the ratio between the largest and smallest radius is\nat most four.\n",
"title": "Placing your Coins on a Shelf"
}
| null | null | null | null | true | null |
9377
| null |
Default
| null | null |
null |
{
"abstract": " The randomized rumor spreading problem generates a big interest in the area\nof distributed algorithms due to its simplicity, robustness and wide range of\napplications. The two most popular communication paradigms used for spreading\nthe rumor are Push and Pull algorithms. The former protocol allows nodes to\nsend the rumor to a randomly selected neighbor at each step, while the latter\nis based on sending a request and downloading the rumor from a randomly\nselected neighbor, provided the neighbor has it. Previous analysis of these\nprotocols assumed that every node could process all such push/pull operations\nwithin a single step, which could be unrealistic in practical situations.\nTherefore we propose a new framework for analysis rumor spreading accommodating\nbuffers, in which a node can process only one push/pull message or push request\nat a time. We develop upper and lower bounds for randomized rumor spreading\ntime in the new framework, and compare the results with analogous in the old\nframework without buffers.\n",
"title": "Randomized Rumor Spreading in Ad Hoc Networks with Buffers"
}
| null | null | null | null | true | null |
9378
| null |
Default
| null | null |
null |
{
"abstract": " Due to severe mathematical modeling and calibration difficulties open-loop\nfeedforward control is mainly employed today for wastewater denitrification,\nwhich is a key ecological issue. In order to improve the resulting poor\nperformances a new model-free control setting and its corresponding\n\"intelligent\" controller are introduced. The pitfall of regulating two output\nvariables via a single input variable is overcome by introducing also an\nopen-loop knowledge-based control deduced from the plant behavior. Several\nconvincing computer simulations are presented and discussed.\n",
"title": "A simple and efficient feedback control strategy for wastewater denitrification"
}
| null | null | null | null | true | null |
9379
| null |
Default
| null | null |
null |
{
"abstract": " We employ the generic three-wave system, with the $\\chi ^{(2)}$ interaction\nbetween two components of the fundamental-frequency (FF) wave and\nsecond-harmonic (SH) one, to consider collisions of truncated Airy waves (TAWs)\nand three-wave solitons in a setting which is not available in other nonlinear\nsystems. The advantage is that the single-wave TAWs, carried by either one of\nthe FF component, are not distorted by the nonlinearity and are stable,\nthree-wave solitons being stable too in the same system. The collision between\nmutually symmetric TAWs, carried by the different FF components, transforms\nthem into a set of solitons, the number of which decreases with the increase of\nthe total power. The TAW absorbs an incident small-power soliton, and a\nhigh-power soliton absorbs the TAW. Between these limits, the collision with an\nincident soliton converts the TAW into two solitons, with a remnant of the TAW\nattached to one of them, or leads to formation of a complex TAW-soliton bound\nstate. At large velocities, the collisions become quasi-elastic.\n",
"title": "The interaction of Airy waves and solitons in the three-wave system"
}
| null | null | null | null | true | null |
9380
| null |
Default
| null | null |
null |
{
"abstract": " This note proposes a penalty criterion for assessing correct score\nforecasting in a soccer match. The penalty is based on hierarchical priorities\nfor such a forecast i.e., i) Win, Draw and Loss exact prediction and ii)\nnormalized Euclidian distance between actual and forecast scores. The procedure\nis illustrated on typical scores, and different alternatives on the penalty\ncomponents are discussed.\n",
"title": "A penalty criterion for score forecasting in soccer"
}
| null | null | null | null | true | null |
9381
| null |
Default
| null | null |
null |
{
"abstract": " Identification of differentially expressed genes (DE-genes) is commonly\nconducted in modern biomedical researches. However, unwanted variation\ninevitably arises during the data collection process, which could make the\ndetection results heavily biased. It is suggested to remove the unwanted\nvariation while keeping the biological variation to ensure a reliable analysis\nresult. Removing Unwanted Variation (RUV) is recently proposed for this purpose\nby the virtue of negative control genes. On the other hand, outliers are\nfrequently appear in modern high-throughput genetic data that can heavily\naffect the performances of RUV and its downstream analysis. In this work, we\npropose a robust RUV-testing procedure via gamma-divergence. The advantages of\nour method are twofold: (1) it does not involve any modeling for the outlier\ndistribution, which is applicable to various situations, (2) it is easy to\nimplement in the sense that its robustness is controlled by a single tuning\nparameter gamma of gamma-divergence, and a data-driven criterion is developed\nto select $\\gamma$. In the Gender Study, our method can successfully remove\nunwanted variation, and is able to identify more DE-genes than conventional\nmethods.\n",
"title": "A robust RUV-testing procedure via gamma-divergence"
}
| null | null | null | null | true | null |
9382
| null |
Default
| null | null |
null |
{
"abstract": " We derive expressions for the finite-sample distribution of the Lasso\nestimator in the context of a linear regression model with normally distributed\nerrors in low as well as in high dimensions by exploiting the structure of the\noptimization problem defining the estimator. In low dimensions we assume full\nrank of the regressor matrix and present expressions for the cumulative\ndistribution function as well as the densities of the absolutely continuous\nparts of the estimator. Additionally, we establish an explicit formula for the\ncorrespondence between the Lasso and the least-squares estimator. We derive\nanalogous results for the distribution in less explicit form in high dimensions\nwhere we make no assumptions on the regressor matrix at all. In this setting,\nwe also investigate the model selection properties of the Lasso and show that\npossibly only a subset of models might be selected by the estimator, completely\nindependently of the observed response vector. Finally, we present a condition\nfor uniqueness of the estimator that is necessary as well as sufficient.\n",
"title": "On the Distribution, Model Selection Properties and Uniqueness of the Lasso Estimator in Low and High Dimensions"
}
| null | null | null | null | true | null |
9383
| null |
Default
| null | null |
null |
{
"abstract": " We present transductive Boltzmann machines (TBMs), which firstly achieve\ntransductive learning of the Gibbs distribution. While exact learning of the\nGibbs distribution is impossible by the family of existing Boltzmann machines\ndue to combinatorial explosion of the sample space, TBMs overcome the problem\nby adaptively constructing the minimum required sample space from data to avoid\nunnecessary generalization. We theoretically provide bias-variance\ndecomposition of the KL divergence in TBMs to analyze its learnability, and\nempirically demonstrate that TBMs are superior to the fully visible Boltzmann\nmachines and popularly used restricted Boltzmann machines in terms of\nefficiency and effectiveness.\n",
"title": "Transductive Boltzmann Machines"
}
| null | null |
[
"Statistics"
] | null | true | null |
9384
| null |
Validated
| null | null |
null |
{
"abstract": " This paper presents a human-robot trust integrated task allocation and motion\nplanning framework for multi-robot systems (MRS) in performing a set of tasks\nconcurrently. A set of task specifications in parallel are conjuncted with MRS\nto synthesize a task allocation automaton. Each transition of the task\nallocation automaton is associated with the total trust value of human in\ncorresponding robots. Here, the human-robot trust model is constructed with a\ndynamic Bayesian network (DBN) by considering individual robot performance,\nsafety coefficient, human cognitive workload and overall evaluation of task\nallocation. Hence, a task allocation path with maximum encoded human-robot\ntrust can be searched based on the current trust value of each robot in the\ntask allocation automaton. Symbolic motion planning (SMP) is implemented for\neach robot after they obtain the sequence of actions. The task allocation path\ncan be intermittently updated with this DBN based trust model. The overall\nstrategy is demonstrated by a simulation with 5 robots and 3 parallel subtask\nautomata.\n",
"title": "Human-Robot Trust Integrated Task Allocation and Symbolic Motion planning for Heterogeneous Multi-robot Systems"
}
| null | null | null | null | true | null |
9385
| null |
Default
| null | null |
null |
{
"abstract": " A statistical test can be seen as a procedure to produce a decision based on\nobserved data, where some decisions consist of rejecting a hypothesis (yielding\na significant result) and some do not, and where one controls the probability\nto make a wrong rejection at some pre-specified significance level. Whereas\ntraditional hypothesis testing involves only two possible decisions (to reject\nor not a null hypothesis), Kaiser's directional two-sided test as well as the\nmore recently introduced Jones and Tukey's testing procedure involve three\npossible decisions to infer on unidimensional parameter. The latter procedure\nassumes that a point null hypothesis is impossible (e.g. that two treatments\ncannot have exactly the same effect), allowing a gain of statistical power.\nThere are however situations where a point hypothesis is indeed plausible, for\nexample when considering hypotheses derived from Einstein's theories. In this\narticle, we introduce a five-decision rule testing procedure, which combines\nthe advantages of the testing procedures of Kaiser (no assumption on a point\nhypothesis being impossible) and of Jones and Tukey (higher power), allowing\nfor a non-negligible (typically 20%) reduction of the sample size needed to\nreach a given statistical power to get a significant result, compared to the\ntraditional approach.\n",
"title": "A five-decision testing procedure to infer on unidimensional parameter"
}
| null | null | null | null | true | null |
9386
| null |
Default
| null | null |
null |
{
"abstract": " We present PubMed 200k RCT, a new dataset based on PubMed for sequential\nsentence classification. The dataset consists of approximately 200,000\nabstracts of randomized controlled trials, totaling 2.3 million sentences. Each\nsentence of each abstract is labeled with their role in the abstract using one\nof the following classes: background, objective, method, result, or conclusion.\nThe purpose of releasing this dataset is twofold. First, the majority of\ndatasets for sequential short-text classification (i.e., classification of\nshort texts that appear in sequences) are small: we hope that releasing a new\nlarge dataset will help develop more accurate algorithms for this task. Second,\nfrom an application perspective, researchers need better tools to efficiently\nskim through the literature. Automatically classifying each sentence in an\nabstract would help researchers read abstracts more efficiently, especially in\nfields where abstracts may be long, such as the medical field.\n",
"title": "PubMed 200k RCT: a Dataset for Sequential Sentence Classification in Medical Abstracts"
}
| null | null | null | null | true | null |
9387
| null |
Default
| null | null |
null |
{
"abstract": " We address the problem of analyzing the radius of convergence of perturbative\nexpansion of non-equilibrium steady states of Lindblad driven spin chains. A\nsimple formal approach is developed for systematically computing the\nperturbative expansion of small driven systems. We consider the paradigmatic\nmodel of an open $XXZ$ spin 1/2 chain with boundary supported ultralocal\nLindblad dissipators and treat two different perturbative cases: (i) expansion\nin system-bath coupling parameter and (ii) expansion in driving (bias)\nparameter. In the first case (i) we find that the radius of convergence quickly\nshrinks with increasing the system size, while in the second case (ii) we find\nthat the convergence radius is always larger than $1$, and in particular it\napproaches $1$ from above as we change the anisotropy from easy plane ($XY$) to\neasy axis (Ising) regime.\n",
"title": "Convergence radius of perturbative Lindblad driven non-equilibrium steady states"
}
| null | null | null | null | true | null |
9388
| null |
Default
| null | null |
null |
{
"abstract": " Customarily, in-plane auxeticity and synclastic bending behavior (i.e.\nout-of-plane auxeticity) are not independent, being the latter a manifestation\nof the former. Basically, this is a feature of three-dimensional bodies. At\nvariance, two-dimensional bodies have more freedom to deform than\nthree-dimensional ones. Here, we exploit this peculiarity and propose a\ntwo-dimensional honeycomb structure with out-of-plane auxetic behavior opposite\nto the in-plane one. With a suitable choice of the lattice constitutive\nparameters, in its continuum description such a structure can achieve the whole\nrange of values for the bending Poisson coefficient, while retaining a\nmembranal Poisson coefficient equal to 1. In particular, this structure can\nreach the extreme values, $-1$ and $+1$, of the bending Poisson coefficient.\nAnalytical calculations are supported by numerical simulations, showing the\naccuracy of the continuum formulas in predicting the response of the discrete\nstructure.\n",
"title": "A 2D metamaterial with auxetic out-of-plane behavior and non-auxetic in-plane behavior"
}
| null | null | null | null | true | null |
9389
| null |
Default
| null | null |
null |
{
"abstract": " A scheme making use of an isolated feedback loop was recently proposed in\n\\cite{GP_} for creating an arbitrary bilinear Hamiltonian interaction between\ntwo multi-mode Linear Quantum Stochastic Systems (LQSSs). In this work we\nexamine the presence of an isolated feedback loop in a general SLH network, and\nderive the modified Hamiltonian of the network due to the presence of the loop.\nIn the case of a bipartite network with an isolated loop running through both\nparts, this results in modified Hamiltonians for each subnetwork, as well as a\nHamiltonian interaction between them. As in the LQSS case, by engineering\nappropriate ports in each subnetwork, we may create desired interactions\nbetween them. Examples are provided that illustrate the general theory.\n",
"title": "Isolated Loops in Quantum Feedback Networks"
}
| null | null | null | null | true | null |
9390
| null |
Default
| null | null |
null |
{
"abstract": " Directional data are constrained to lie on the unit sphere of~$\\mathbb{R}^q$\nfor some~$q\\geq 2$. To address the lack of a natural ordering for such data,\ndepth functions have been defined on spheres. However, the depths available\neither lack flexibility or are so computationally expensive that they can only\nbe used for very small dimensions~$q$. In this work, we improve on this by\nintroducing a class of distance-based depths for directional data. Irrespective\nof the distance adopted, these depths can easily be computed in high dimensions\ntoo. We derive the main structural properties of the proposed depths and study\nhow they depend on the distance used. We discuss the asymptotic and robustness\nproperties of the corresponding deepest points. We show the practical relevance\nof the proposed depths in two applications, related to (i) spherical location\nestimation and (ii) supervised classification. For both problems, we show\nthrough simulation studies that distance-based depths have strong advantages\nover their competitors.\n",
"title": "Distance-based Depths for Directional Data"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
9391
| null |
Validated
| null | null |
null |
{
"abstract": " Bose-Einstein condensates (BECs) confined in a two-dimensional (2D) harmonic\ntrap are known to possess a hidden 2D Schrödinger symmetry, that is, the\nSchrödinger symmetry modified by a trapping potential. Spontaneous breaking\nof this symmetry gives rise to a breathing motion of the BEC, whose oscillation\nfrequency is robustly determined by the strength of the harmonic trap. In this\npaper, we demonstrate that the concept of the 2D Schrödinger symmetry can be\napplied to predict the nature of three dimensional (3D) collective modes\npropagating along a condensate confined in an elongated trap. We find three\nkinds of collective modes whose existence is robustly ensured by the\nSchrödinger symmetry, which are physically interpreted as one breather mode\nand two Kelvin-ripple complex modes, i.e., composite modes in which the vortex\ncore and the condensate surface oscillate interactively. We provide analytical\nexpressions for the dispersion relations (energy-momentum relation) of these\nmodes using the Bogoliubov theory [D. A. Takahashi and M. Nitta, Ann. Phys.\n354, 101 (2015)]. Furthermore, we point out that these modes can be interpreted\nas \"quasi-massive-Nambu-Goldstone (NG) modes\", that is, they have the\nproperties of both quasi-NG and massive NG modes: quasi-NG modes appear when a\nsymmetry of a part of a Lagrangian, which is not a symmetry of full a\nLagrangian, is spontaneously broken, while massive NG modes appear when a\nmodified symmetry is spontaneously broken.\n",
"title": "Two-dimensional Schrödinger symmetry and three-dimensional breathers and Kelvin-ripple complexes as quasi-massive-Nambu-Goldstone modes"
}
| null | null |
[
"Physics"
] | null | true | null |
9392
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we construct two groupoids from morphisms of groupoids, with\none from a categorical viewpoint and the other from a geometric viewpoint. We\nshow that for each pair of groupoids, the two kinds of groupoids of morphisms\nare equivalent. Then we study the automorphism groupoid of a groupoid.\n",
"title": "Groupoid of morphisms of groupoids"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9393
| null |
Validated
| null | null |
null |
{
"abstract": " Stellar clusters form by gravitational collapse of turbulent molecular\nclouds, with up to several thousand stars per cluster. They are thought to be\nthe birthplace of most stars and therefore play an important role in our\nunderstanding of star formation, a fundamental problem in astrophysics. The\ninitial conditions of the molecular cloud establish its dynamical history until\nthe stellar cluster is born. However, the evolution of the cloud's angular\nmomentum during cluster formation is not well understood. Current observations\nhave suggested that turbulence scrambles the angular momentum of the\ncluster-forming cloud, preventing spin alignment amongst stars within a\ncluster. Here we use asteroseismology to measure the inclination angles of spin\naxes in 48 stars from the two old open clusters NGC~6791 and NGC~6819. The\nstars within each cluster show strong alignment. Three-dimensional\nhydrodynamical simulations of proto-cluster formation show that at least 50 %\nof the initial proto-cluster kinetic energy has to be rotational in order to\nobtain strong stellar-spin alignment within a cluster. Our result indicates\nthat the global angular momentum of the cluster-forming clouds was efficiently\ntransferred to each star and that its imprint has survived after several\ngigayears since the clusters formed.\n",
"title": "Spin alignment of stars in old open clusters"
}
| null | null |
[
"Physics"
] | null | true | null |
9394
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we prove global well-posedness of the critical surface\nquasigeostrophic equation on the two dimensional sphere building on some\nearlier work of the authors. The proof relies on an improving of the previously\nknown pointwise inequality for fractional laplacians as in the work of\nConstantin and Vicol for the euclidean setting.\n",
"title": "Global well-posedness of critical surface quasigeostrophic equation on the sphere"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9395
| null |
Validated
| null | null |
null |
{
"abstract": " Understanding and characterizing the subspaces of adversarial examples aid in\nstudying the robustness of deep neural networks (DNNs) to adversarial\nperturbations. Very recently, Ma et al. (ICLR 2018) proposed to use local\nintrinsic dimensionality (LID) in layer-wise hidden representations of DNNs to\nstudy adversarial subspaces. It was demonstrated that LID can be used to\ncharacterize the adversarial subspaces associated with different attack\nmethods, e.g., the Carlini and Wagner's (C&W) attack and the fast gradient sign\nattack.\nIn this paper, we use MNIST and CIFAR-10 to conduct two new sets of\nexperiments that are absent in existing LID analysis and report the limitation\nof LID in characterizing the corresponding adversarial subspaces, which are (i)\noblivious attacks and LID analysis using adversarial examples with different\nconfidence levels; and (ii) black-box transfer attacks. For (i), we find that\nthe performance of LID is very sensitive to the confidence parameter deployed\nby an attack, and the LID learned from ensembles of adversarial examples with\nvarying confidence levels surprisingly gives poor performance. For (ii), we\nfind that when adversarial examples are crafted from another DNN model, LID is\nineffective in characterizing their adversarial subspaces. These two findings\ntogether suggest the limited capability of LID in characterizing the subspaces\nof adversarial examples.\n",
"title": "On the Limitation of Local Intrinsic Dimensionality for Characterizing the Subspaces of Adversarial Examples"
}
| null | null | null | null | true | null |
9396
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we construct a properly embedded holomorphic disc in the unit\nball $\\mathbb{B}^2$ of $\\mathbb{C}^2$ having a surprising combination of\nproperties: on the one hand, it has finite area and hence is the zero set of a\nbounded holomorphic function on $\\mathbb{B}^2$; on the other hand, its boundary\ncurve is everywhere dense in the sphere $b\\mathbb{B}^2$.\n",
"title": "A properly embedded holomorphic disc in the ball with finite area and dense boundary curve"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9397
| null |
Validated
| null | null |
null |
{
"abstract": " Given a 0-dimensional scheme in a projective space $\\mathbb{P}^n$ over a\nfield $K$, we study the Kähler differential algebra $\\Omega_{R/K}$ of its\nhomogeneous coordinate ring $R$. Using explicit presentations of the modules\n$\\Omega^m_{R/K}$ of Kähler differential $m$-forms, we determine many values\nof their Hilbert functions explicitly and bound their Hilbert polynomials and\nregularity indices. Detailed results are obtained for subschemes of\n$\\mathbb{P}^1$, fat point schemes, and subschemes of $\\mathbb{P}^2$ supported\non a conic.\n",
"title": "Kähler differential algebras for 0-dimensional schemes"
}
| null | null |
[
"Mathematics"
] | null | true | null |
9398
| null |
Validated
| null | null |
null |
{
"abstract": " DFT is used throughout nanoscience, especially when modeling spin-dependent\nproperties that are important in spintronics. But standard quantum chemical\nmethods (both CCSD(T) and self-consistent semilocal density functional\ncalculations) fail badly for the spin adiabatic energy difference in Fe(II)\nspin-crossover complexes. We show that all-electron fixed-node diffusion Monte\nCarlo can be converged at significant computational cost, and that the B3LYP\nsingle-determinant has sufficiently accurate nodes, providing benchmarks for\nthese systems. We also find that density-corrected DFT, using Hartree-Fock\ndensities (HF-DFT), greatly improves accuracy and reduces dependence on\napproximations for these calculations. The small gap in the self-consistent DFT\ncalculations for the high-spin state is consistent with this. For the spin\nadiabatic energy differences in these complexes, HF-DFT is both accurate and\nreliable, and we make a strong prediction for the Fe-Porphyrin complex. The\n\"parameter-dilemma\" of needing different amounts of mixing for different\nproperties is eliminated by HF-DFT.\n",
"title": "Benchmarks and reliable DFT results for spin-crossover complexes"
}
| null | null | null | null | true | null |
9399
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the birth and diffusion of lexical innovations in a large\ndataset of online social communities. We build on sociolinguistic theories and\nfocus on the relation between the spread of a novel term and the social role of\nthe individuals who use it, uncovering characteristics of innovators and\nadopters. Finally, we perform a prediction task that allows us to anticipate\nwhether an innovation will successfully spread within a community.\n",
"title": "The Road to Success: Assessing the Fate of Linguistic Innovations in Online Communities"
}
| null | null |
[
"Computer Science"
] | null | true | null |
9400
| null |
Validated
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.