arXiv subjects

Yi Yang

Publications and source records attributed to Yi Yang.

At least 1,099 records · Page 61Linked to original sources

arXiv2016

Toward Socially-Infused Information Extraction: Embedding Authors, Mentions, and Entities

Entity linking is the task of identifying mentions of entities in text, and linking them to entries in a knowledge base. This task is especially difficult in microblogs, as there is little additional text to provide disambiguating context; rather, authors rely on an implicit common ground of shared knowledge with their readers. In this paper, we attempt to capture some of this implicit context by exploiting the social network structure in microblogs. We build on the theory of homophily, which implies that socially linked individuals share interests, and are therefore likely to mention the same sorts of entities. We implement this idea by encoding authors, mentions, and entities in a continuous vector space, which is constructed so that socially-connected authors have similar vector representations. These vectors are incorporated into a neural structured prediction model, which captures structural constraints that are inherent in the entity linking task. Together, these design decisions yield F1 improvements of 1%-5% on benchmark datasets, as compared to the previous state-of-the-art.

cs.CL↗

arXiv2016

The Lauricella Functions and Exact String Scattering Amplitudes

We discover that the 26D open bosonic string scattering amplitudes (SSA) of three tachyons and one arbitrary string state can be expressed in terms of the D-type Lauricella functions with associated SL(K+3,C) symmetry. As a result, SSA and symmetries or relations among SSA of different string states at various limits calculated previously can be rederived. These include the linear relations first conjectured by Gross [1-3] and later corrected and proved in [4-9] in the hard scattering limit, the recurrence relations in the Regge scattering limit with associated SL(5,C) symmetry [19-21] and the extended recurrence relations in the nonrelativistic scattering limit with associated SL(4,C) symmetry [24] discovered recently. Finally, as an application, we calculate a new recurrence relation of SSA which is valid for all energies.

hep-th↗

arXiv2016

ASASSN-15lh: A Superluminous Ultraviolet Rebrightening Observed by Swift and Hubble

We present and discuss ultraviolet and optical photometry from the Ultraviolet/Optical Telescope and X-ray limits from the X-Ray Telescope on Swift and imaging polarimetry and ultraviolet/optical spectroscopy with the Hubble Space Telescope of ASASSN-15lh. It has been classified as a hydrogen-poor superluminous supernova (SLSN I) more luminous than any other supernova observed. ASASSN-15lh is not detected in the X-rays in individual or coadded observations. From the polarimetry we determine that the explosion was only mildly asymmetric. We find the flux of ASASSN-15lh to increase strongly into the ultraviolet, with a ultraviolet luminosity a hundred times greater than the hydrogen-rich, ultraviolet-bright SLSN II SN 2008es. We find objects as bright as ASASSN-15lh are easily detectable beyond redshifts of ~4 with the single-visit depths planned for the Large Synoptic Survey Telescope. Deep near-infrared surveys could detect such objects past a redshift of ~20 enabling a probe of the earliest star formation. A late rebrightening -- most prominent at shorter wavelengths -- is seen about two months after the peak brightness, which is itself as bright as a superluminous supernova. The ultraviolet spectra during the rebrightening are dominated by the continuum without the broad absorption or emission lines seen in SLSNe or tidal disruption events and the early optical spectra of ASASSN-15lh. Our spectra show no strong hydrogen emission, showing only LyA absorption near the redshift previously found by optical absorption lines of the presumed host. The properties of ASASSN-15lh are extreme when compared to either SLSNe or tidal disruption events.

astro-ph.HE↗

arXiv2016

Sparsity Oriented Importance Learning for High-dimensional Linear Regression

With now well-recognized non-negligible model selection uncertainty, data analysts should no longer be satisfied with the output of a single final model from a model selection process, regardless of its sophistication. To improve reliability and reproducibility in model choice, one constructive approach is to make good use of a sound variable importance measure. Although interesting importance measures are available and increasingly used in data analysis, little theoretical justification has been done. In this paper, we propose a new variable importance measure, sparsity oriented importance learning (SOIL), for high-dimensional regression from a sparse linear modeling perspective by taking into account the variable selection uncertainty via the use of a sensible model weighting. The SOIL method is theoretically shown to have the inclusion/exclusion property: When the model weights are properly around the true model, the SOIL importance can well separate the variables in the true model from the rest. In particular, even if the signal is weak, SOIL rarely gives variables not in the true model significantly higher important values than those in the true model. Extensive simulations in several illustrative settings and real data examples with guided simulations show desirable properties of the SOIL importance in contrast to other importance measures.

stat.ME↗

arXiv2016

Strategies for Searching Video Content with Text Queries or Video Examples

The large number of user-generated videos uploaded on to the Internet everyday has led to many commercial video search engines, which mainly rely on text metadata for search. However, metadata is often lacking for user-generated videos, thus these videos are unsearchable by current search engines. Therefore, content-based video retrieval (CBVR) tackles this metadata-scarcity problem by directly analyzing the visual and audio streams of each video. CBVR encompasses multiple research topics, including low-level feature design, feature fusion, semantic detector training and video search/reranking. We present novel strategies in these topics to enhance CBVR in both accuracy and speed under different query inputs, including pure textual queries and query by video examples. Our proposed strategies have been incorporated into our submission for the TRECVID 2014 Multimedia Event Detection evaluation, where our system outperformed other submissions in both text queries and video example queries, thus demonstrating the effectiveness of our proposed approaches.

cs.IR↗

arXiv2016

Optically Thin Metallic Films for High-radiative-efficiency Plasmonics

Plasmonics enables deep-subwavelength concentration of light and has become important for fundamental studies as well as real-life applications. Two major existing platforms of plasmonics are metallic nanoparticles and metallic films. Metallic nanoparticles allow efficient coupling to far field radiation, yet their synthesis typically leads to poor material quality. Metallic films offer substantially higher quality materials, but their coupling to radiation is typically jeopardized due to the large momentum mismatch with free space. Here, we propose and theoretically investigate optically thin metallic films as an ideal platform for high-radiative-efficiency plasmonics. For far-field scattering, adding a thin high-quality metallic substrate enables a higher quality factor while maintaining the localization and tunability that the nanoparticle provides. For near-field spontaneous emission, a thin metallic substrate, of high quality or not, greatly improves the field overlap between the emitter environment and propagating surface plasmons, enabling high-Purcell (total enhancement > $10^4$), high-quantum-yield (> 50 %) spontaneous emission, even as the gap size vanishes (3$\sim$5 nm). The enhancement has almost spatially independent efficiency and does not suffer from quenching effects that commonly exist in previous structures.

physics.optics↗

arXiv2016

Probing Neutrino Mass Hierarchy by Comparing the Charged-Current and Neutral-Current Interaction Rates of Supernova Neutrinos

The neutrino mass hierarchy is one of the neutrino fundamental properties yet to be determined. We introduce a method to determine neutrino mass hierarchy by comparing the interaction rate of neutral current (NC) interactions, $ν(\hatν) + p\rightarrowν(\hatν) + p$, and inverse beta decays (IBD), $\barν_e + p\rightarrow n + e^+$, of supernova neutrinos in scintillation detectors. Neutrino flavor conversions inside the supernova are sensitive to neutrino mass hierarchy. Due to Mikheyev-Smirnov-Wolfenstein effects, the full swapping of $\barν_e$ flux with the $\barν_x$ ($x=μ,~τ$) one occurs in the inverted hierarchy, while such a swapping does not occur in the normal hierarchy. As a result, more high energy IBD events occur in the detector for the inverted hierarchy than the high energy IBD events in the normal hierarchy. By comparing IBD interaction rate with the mass hierarchy independent NC interaction rate, one can determine the neutrino mass hierarchy.

hep-ph↗

arXiv2016

Attention to Scale: Scale-aware Semantic Image Segmentation

Incorporating multi-scale features in fully convolutional neural networks (FCNs) has been a key element to achieving state-of-the-art performance on semantic image segmentation. One common way to extract multi-scale features is to feed multiple resized input images to a shared deep network and then merge the resulting features for pixelwise classification. In this work, we propose an attention mechanism that learns to softly weight the multi-scale features at each pixel location. We adapt a state-of-the-art semantic image segmentation model, which we jointly train with multi-scale input images and the attention model. The proposed attention model not only outperforms average- and max-pooling, but allows us to diagnostically visualize the importance of features at different positions and scales. Moreover, we show that adding extra supervision to the output at each scale is essential to achieving excellent performance when merging multi-scale features. We demonstrate the effectiveness of our model with extensive experiments on three challenging datasets, including PASCAL-Person-Part, PASCAL VOC 2012 and a subset of MS-COCO 2014.

cs.CV↗

arXiv2016

Fluid/Gravity Correspondence with Scalar Field and Electromagnetic Field

We consider fluid/gravity correspondence in a general rotating black hole background with scalar and electromagnetic fields. Using the method of Petrov-like boundary condition, we show that the scalar and the electromagnetic fields contribute external forces to the dual Navier-Stokes equation and the rotation of black hole induces the Coriolis force.

hep-th↗

arXiv2016

The String BCJ Relations Revisited and Extended Recurrence relations of Nonrelativistic String Scattering Amplitudes

We review and extend high energy string BCJ relations in both the fixed angle and Regge regimes. We then give an explicit proof of four point string BCJ relations for all energy. This calculation provides an alternative proof of the one based on monodromy of integration in string amplitude calculation. In addition, we calculate both s-t and t-u channel nonrelativistic low energy string scattering amplitudes of three tachyons and one leading trojectory string state at arbitrary mass levels. We discover that the mass and spin dependent nonrelativistic string BCJ relations can be expressed in terms of Gauss hypergeometry functions. As an application, for each fixed mass level N, we derive extended recurrence relations among nonrelativistic low energy string scattering amplitudes of string states with different spins and different channels.

hep-th↗

arXiv2016

Insurance Premium Prediction via Gradient Tree-Boosted Tweedie Compound Poisson Models

The Tweedie GLM is a widely used method for predicting insurance premiums. However, the structure of the logarithmic mean is restricted to a linear form in the Tweedie GLM, which can be too rigid for many applications. As a better alternative, we propose a gradient tree-boosting algorithm and apply it to Tweedie compound Poisson models for pure premiums. We use a profile likelihood approach to estimate the index and dispersion parameters. Our method is capable of fitting a flexible nonlinear Tweedie model and capturing complex interactions among predictors. A simulation study confirms the excellent prediction performance of our method. As an application, we apply our method to an auto insurance claim data and show that the new method is superior to the existing methods in the sense that it generates more accurate premium predictions, thus helping solve the adverse selection issue. We have implemented our method in a user-friendly R package that also includes a nice visualization tool for interpreting the fitted model.

stat.ME↗

arXiv2016

CNN-RNN: A Unified Framework for Multi-label Image Classification

While deep convolutional neural networks (CNNs) have shown a great success in single-label image classification, it is important to note that real world images generally contain multiple labels, which could correspond to different objects, scenes, actions and attributes in an image. Traditional approaches to multi-label image classification learn independent classifiers for each category and employ ranking or thresholding on the classification results. These techniques, although working well, fail to explicitly exploit the label dependencies in an image. In this paper, we utilize recurrent neural networks (RNNs) to address this problem. Combined with CNNs, the proposed CNN-RNN framework learns a joint image-label embedding to characterize the semantic label dependency as well as the image-label relevance, and it can be trained end-to-end from scratch to integrate both information in a unified framework. Experimental results on public benchmark datasets demonstrate that the proposed architecture achieves better performance than the state-of-the-art multi-label classification model

cs.CV↗

arXiv2016

Video Paragraph Captioning Using Hierarchical Recurrent Neural Networks

We present an approach that exploits hierarchical Recurrent Neural Networks (RNNs) to tackle the video captioning problem, i.e., generating one or multiple sentences to describe a realistic video. Our hierarchical framework contains a sentence generator and a paragraph generator. The sentence generator produces one simple short sentence that describes a specific short video interval. It exploits both temporal- and spatial-attention mechanisms to selectively focus on visual elements during generation. The paragraph generator captures the inter-sentence dependency by taking as input the sentential embedding produced by the sentence generator, combining it with the paragraph history, and outputting the new initial state for the sentence generator. We evaluate our approach on two large-scale benchmark datasets: YouTubeClips and TACoS-MultiLevel. The experiments demonstrate that our approach significantly outperforms the current state-of-the-art methods with BLEU@4 scores 0.499 and 0.305 respectively.

cs.CV↗

arXiv2016

Part-of-Speech Tagging for Historical English

As more historical texts are digitized, there is interest in applying natural language processing tools to these archives. However, the performance of these tools is often unsatisfactory, due to language change and genre differences. Spelling normalization heuristics are the dominant solution for dealing with historical texts, but this approach fails to account for changes in usage and vocabulary. In this empirical paper, we assess the capability of domain adaptation techniques to cope with historical texts, focusing on the classic benchmark task of part-of-speech tagging. We evaluate several domain adaptation methods on the task of tagging Early Modern English and Modern British English texts in the Penn Corpora of Historical English. We demonstrate that the Feature Embedding method for unsupervised domain adaptation outperforms word embeddings and Brown clusters, showing the importance of embedding the entire feature space, rather than just individual words. Feature Embeddings also give better performance than spelling normalization, but the combination of the two methods is better still, yielding a 5% raw improvement in tagging accuracy on Early Modern English texts.

cs.CL↗

arXiv2016

The Exact SL(K+3,C) Symmetry of String Scattering Amplitudes

We discover that the 26D open bosonic string scattering amplitudes (SSA) of three tachyons and one arbitrary string state can be expressed in terms of the D-type Lauricella functions with associated SL(K+3,C) symmetry. As a result, SSA and symmetries or relations among SSA of different string states at various limits calculated previously can be rederived. These include the linear relations conjectured by Gross [1-3] and proved in [4-9] in the hard scattering limit, the recurrence relations in the Regge scattering limit [14-16] and the extended recurrence relations in the nonrelativistic scattering limit [19] discovered recently. Finally, as an application, we calculate a new recurrence relation of SSA which is valid for all energies.

hep-th↗

arXiv2016

Dynamic Concept Composition for Zero-Example Event Detection

In this paper, we focus on automatically detecting events in unconstrained videos without the use of any visual training exemplars. In principle, zero-shot learning makes it possible to train an event detection model based on the assumption that events (e.g. \emph{birthday party}) can be described by multiple mid-level semantic concepts (e.g. "blowing candle", "birthday cake"). Towards this goal, we first pre-train a bundle of concept classifiers using data from other sources. Then we evaluate the semantic correlation of each concept \wrt the event of interest and pick up the relevant concept classifiers, which are applied on all test videos to get multiple prediction score vectors. While most existing systems combine the predictions of the concept classifiers with fixed weights, we propose to learn the optimal weights of the concept classifiers for each testing video by exploring a set of online available videos with free-form text descriptions of their content. To validate the effectiveness of the proposed approach, we have conducted extensive experiments on the latest TRECVID MEDTest 2014, MEDTest 2013 and CCV dataset. The experimental results confirm the superiority of the proposed approach.

cs.CV↗

arXiv2016

Confinement-Deconfinment Phase Transition for Heavy Quarks

We study confinement-deconfinement phase transition for heavy quarks in a bottom-up holographic QCD model. We consider a black hole background in an Einstein-Maxwell-scalar system and add probe open strings to the background. Combining the various configurations of the open strings and the phase structure of the black hole background itself, we obtain the confinement-deconfinement phase diagram for heavy quarks in the holographic QCD model.

hep-th↗

arXiv2015

Extremal RN/CFT in Both Hands Revisited

We study RN/CFT correspondence for four dimensional extremal Reissner-Nordstrom black hole. We uplift the 4d RN black hole to a 5d rotating black hole and make a geometric regularization of the 5d space-time. Both hands central charges are obtained correctly at the same time by Brown-Henneaux technique.

hep-th↗