arXiv subjects

Yue Zhang

Publications and source records attributed to Yue Zhang.

At least 739 records · Page 41Linked to original sources

arXiv2019

Who Blames Whom in a Crisis? Detecting Blame Ties from News Articles Using Neural Networks

Blame games tend to follow major disruptions, be they financial crises, natural disasters or terrorist attacks. To study how the blame game evolves and shapes the dominant crisis narratives is of great significance, as sense-making processes can affect regulatory outcomes, social hierarchies, and cultural norms. However, it takes tremendous time and efforts for social scientists to manually examine each relevant news article and extract the blame ties (A blames B). In this study, we define a new task, Blame Tie Extraction, and construct a new dataset related to the United States financial crisis (2007-2010) from The New York Times, The Wall Street Journal and USA Today. We build a Bi-directional Long Short-Term Memory (BiLSTM) network for contexts where the entities appear in and it learns to automatically extract such blame ties at the document level. Leveraging the large unsupervised model such as GloVe and ELMo, our best model achieves an F1 score of 70% on the test set for blame tie extraction, making it a useful tool for social scientists to extract blame ties more efficiently.

cs.CL↗

arXiv2019

Inverse Compton Scattering Spectra of Gamma-Ray Burst Prompt Emission

Although the physical origin of gamma-ray burst (GRB) prompt emission is still controversial, synchrotron radiation from accelerated electrons is a promising mechanism. It is believed that electrons are accelerated continuously by ultra-relativistic shocks or magnetic reconnections. At the same time, these electrons will be cooled via several processes (mainly adiabatic expansion, synchrotron radiation and inverse Compton scattering (ICS)), which regulate the distribution of electrons. An extra high-energy spectrum component is expected to be induced by ICS. However, the gamma-gamma annihilation effect may eliminate the high-energy photons and prevent the observers from distinguishing the extra component. We perform numerical calculations by taking these effects into account and discuss whether the extra ICS component could be observed. By exploring the plausible parameter space for relevant quantities of the GRB ejecta, we present the electron distributions and the corresponding spectra of synchrotron radiation and ICS. It is found that the extra component is observable only for ejecta with a rather large bulk Lorentz factor. A large Lorentz factor means the adiabatic expansion is the leading process in the electron cooling procedure, which makes the low-energy electron distribution spectrum to be relatively hard. Therefore the ICS component is more likely to be detected for GRBs that have a hard low-energy photon spectrum.

astro-ph.HE↗

arXiv2019

Reverberation mapping of narrow-line Seyfert 1 galaxy I Zwicky 1: black hole mass

We report results of the first reverberation mapping campaign of I Zwicky 1 during $2014$-$2016$, which showed unambiguous reverberations of the broad H$β$ line emission to the varying optical continuum. From analysis using several methods, we obtain a reverberation lag of $τ_{\rm Hβ}=37.2^{+4.5}_{-4.9}\,$ days. Taking a virial factor of $f_{_{\rm BLR}}=1$, we find a black hole mass of $M_{\bullet}=9.30_{-1.38}^{+1.26}\times 10^6 M_{\odot}$ from the mean spectra. The accretion rate is estimated to be $203.9_{-65.8}^{+61.0}\,L_{\rm Edd}c^{-2}$, suggesting a super-Eddington accretor, where $L_{\rm Edd}$ is the Eddington luminosity and $c$ is the speed of light. By decomposing {\it Hubble Space Telescope} images, we find that the stellar mass of the bulge of its host galaxy is $\log (M_{\rm bulge}/M_{\odot}) = \rm 10.92\pm 0.07$. This leads to a black hole to bulge mass ratio of $\sim 10^{-4}$, which is significantly smaller than that of classical bulges and elliptical galaxies. After subtracting the host contamination from the observed luminosity, we find that I Zw 1 follows the empirical $R_{\rm BLR}\propto L_{5100}^{1/2}$ relation.

astro-ph.GA↗

arXiv2019

A Training-free, One-shot Detection Framework For Geospatial Objects In Remote Sensing Images

Deep learning based object detection has achieved great success. However, these supervised learning methods are data-hungry and time-consuming. This restriction makes them unsuitable for limited data and urgent tasks, especially in the applications of remote sensing. Inspired by the ability of humans to quickly learn new visual concepts from very few examples, we propose a training-free, one-shot geospatial object detection framework for remote sensing images. It consists of (1) a feature extractor with remote sensing domain knowledge, (2) a multi-level feature fusion method, (3) a novel similarity metric method, and (4) a 2-stage object detection pipeline. Experiments on sewage treatment plant and airport detections show that proposed method has achieved a certain effect. Our method can serve as a baseline for training-free, one-shot geospatial object detection.

cs.CV↗

arXiv2019

Improving Cross-Domain Chinese Word Segmentation with Word Embeddings

Cross-domain Chinese Word Segmentation (CWS) remains a challenge despite recent progress in neural-based CWS. The limited amount of annotated data in the target domain has been the key obstacle to a satisfactory performance. In this paper, we propose a semi-supervised word-based approach to improving cross-domain CWS given a baseline segmenter. Particularly, our model only deploys word embeddings trained on raw text in the target domain, discarding complex hand-crafted features and domain-specific dictionaries. Innovative subsampling and negative sampling methods are proposed to derive word embeddings optimized for CWS. We conduct experiments on five datasets in special domains, covering domains in novels, medicine, and patent. Results show that our model can obviously improve cross-domain CWS, especially in the segmentation of domain-specific noun entities. The word F-measure increases by over 3.0% on four datasets, outperforming state-of-the-art semi-supervised and unsupervised cross-domain CWS approaches with a large margin. We make our code and data available on Github.

cs.CL↗

arXiv2019

Decentralized Optimal Control of Connected Automated Vehicles at Signal-Free Intersections Including Comfort-Constrained Turns and Safety Guarantees

We extend earlier work for optimally controlling Connected Automated Vehicles (CAVs) crossing a signal-free intersection by including all possible turns taken so as to optimize a passenger comfort metric along with energy and travel time minimization. We show that it is possible to achieve this goal in a decentralized manner with each CAV solving an optimal control problem, and derive explicit solutions that guarantee collision avoidance and safe distance constraints within a control zone. We investigate the associated tradeoffs between minimizing energy and vehicle travel time, as well as the passenger comfort metric and include extensive simulations to illustrate this framework.

math.OC↗

arXiv2019

Cyber-Physical Simulation Platform for Security Assessment of Transactive Energy Systems

Transactive energy systems (TES) are emerging as a transformative solution for the problems that distribution system operators face due to an increase in the use of distributed energy resources and rapid growth in scalability of managing active distribution system (ADS). On the one hand, these changes pose a decentralized power system control problem, requiring strategic control to maintain reliability and resiliency for the community and for the utility. On the other hand, they require robust financial markets while allowing participation from diverse prosumers. To support the computing and flexibility requirements of TES while preserving privacy and security, distributed software platforms are required. In this paper, we enable the study and analysis of security concerns by developing Transactive Energy Security Simulation Testbed (TESST), a TES testbed for simulating various cyber attacks. In this work, the testbed is used for TES simulation with centralized clearing market, highlighting weaknesses in a centralized system. Additionally, we present a blockchain enabled decentralized market solution supported by distributed computing for TES, which on one hand can alleviate some of the problems that we identify, but on the other hand, may introduce newer issues. Future study of these differing paradigms is necessary and will continue as we develop our security simulation testbed.

eess.SY↗

arXiv2019

Semantic Neural Machine Translation using AMR

It is intuitive that semantic representations can be useful for machine translation, mainly because they can help in enforcing meaning preservation and handling data sparsity (many sentences correspond to one meaning) of machine translation models. On the other hand, little work has been done on leveraging semantics for neural machine translation (NMT). In this work, we study the usefulness of AMR (short for abstract meaning representation) on NMT. Experiments on a standard English-to-German dataset show that incorporating AMR as additional knowledge can significantly improve a strong attention-based sequence-to-sequence neural translation model.

cs.CL↗

arXiv2019

Highly Efficient Follicular Segmentation in Thyroid Cytopathological Whole Slide Image

In this paper, we propose a novel method for highly efficient follicular segmentation of thyroid cytopathological WSIs. Firstly, we propose a hybrid segmentation architecture, which integrates a classifier into Deeplab V3 by adding a branch. A large amount of the WSI segmentation time is saved by skipping the irrelevant areas using the classification branch. Secondly, we merge the low scale fine features into the original atrous spatial pyramid pooling (ASPP) in Deeplab V3 to accurately represent the details in cytopathological images. Thirdly, our hybrid model is trained by a criterion-oriented adaptive loss function, which leads the model converging much faster. Experimental results on a collection of thyroid patches demonstrate that the proposed model reaches 80.9% on the segmentation accuracy. Besides, 93% time is reduced for the WSI segmentation by using our proposed method, and the WSI-level accuracy achieves 53.4%.

cs.CV↗

arXiv2019

Recycled Two-Stage Estimation in Nonlinear Mixed Effects Regression Models

We consider a re-sampling scheme for estimation of the population parameters in the mixed effects nonlinear regression models of the type use for example in clinical pharmacokinetics, say. We provide an estimation procedure which {\it recycles}, via random weighting, the relevant two-stage parameters estimates to construct consistent estimates of the sampling distribution of the various estimates. We establish the asymptotic consistency and asymptotic normality of the resampled estimates and demonstrate the applicability of the {\it recycling} approach in a small simulation study and via example.

math.ST↗

arXiv2019

Super-Eddington accretion onto the Galactic ultraluminous X-ray pulsar Swift J0243.6+6124

We report on the spectral behavior of the first Galactic ultraluminous X-ray pulsar Swift J0243.6+6124 with NuSTAR observations during its 2017-2018 outburst. At sub-Eddington levels, the source spectrum is characterized by three emission components, respectively from the accretion column, the hot spot, and a broad iron line emission region. When the source is above the Eddington limit, the hot spot temperature increases and the spectrum features two more blackbody components. One blackbody component has a radius of 10-20 km and is likely originated from the top of the accretion column. The other one saturates at a blackbody luminosity of (1 - 2)*10^38 erg/s, coincident with the Eddington limit of a neutron star. This is well consistent with the scenario that super-Eddington accretion onto compact objects will power optically-thick outflows and indicates an accretion rate 60-80 times the critical value. This suggests that super-Eddington accretion onto magnetized systems can also power massive winds. At super-Eddington levels, the iron line becomes more significant and blueshifted, and is argued to be associated with the ultrafast wind in the central funnel or jets. This source, if located in external galaxies, will appear like other ultraluminous pulsars.

astro-ph.HE↗

arXiv2019

Prostate segmentation using Z-net

In this paper, we proposed a novel architecture of convolutional neural network (CNN), namely Z-net, for segmenting prostate from magnetic resonance images (MRIs). In the proposed Z-net, 5 pairs of Z-block and decoder Z-block with different sizes and numbers of feature maps were assembled in a way similar to that of U-net. The proposed architecture can capture more multi-level features by using concatenation and dense connection. A total of 45 training images were used to train the proposed Z-net and the evaluations were conducted qualitatively on 5 validation images and quantitatively on 30 testing images. In addition, three approaches including pad and cut, 2D resize, and 3D resize for uniforming the size of samples were evaluated and compared. The experimental results demonstrated that the 2D resize is the most suitable approach for the proposed Z-net. Compared to the other two classical CNN architectures, the proposed method was observed with superior performance for segmenting prostate.

eess.IV↗

arXiv2019

Mono-Neutrino at DUNE: New Signals From Neutrinophilic Thermal Dark Matter

We introduce the mono-neutrino signal at neutrino detectors as a smoking gun of sub-GeV scale dark matter candidates that mainly interact with standard model neutrinos. In a mono-neutrino process, invisible particles, either dark matter themselves or the mediator particle, are radiated off a neutrino when it undergoes the charged-current weak interaction. The associated signals include a missing transverse momentum with respect to the incoming neutrino beam direction and the production of wrong-sign charged leptons. We demonstrate the potential leading role of the future DUNE experiment, using its proposed liquid and gas argon near detectors, in probing these new signals and the thermal origins of neutrinophilic dark matter.

hep-ph↗

arXiv2019

Wider Channel Attention Network for Remote Sensing Image Super-resolution

Recently, deep convolutional neural networks (CNNs) have obtained promising results in image processing tasks including super-resolution (SR). However, most CNN-based SR methods treat low-resolution (LR) inputs and features equally across channels, rarely notice the loss of information flow caused by the activation function and fail to leverage the representation ability of CNNs. In this letter, we propose a novel single-image super-resolution (SISR) algorithm named Wider Channel Attention Network (WCAN) for remote sensing images. Firstly, the channel attention mechanism is used to adaptively recalibrate the importance of each channel at the middle of the wider attention block (WAB). Secondly, we propose the Local Memory Connection (LMC) to enhance the information flow. Finally, the features within each WAB are fused to take advantage of the network's representation capability and further improve information and gradient flow. Analytic experiments on a public remote sensing data set (UC Merced) show that our WCAN achieves better accuracy and visual improvements against most state-of-the-art methods.

cs.CV↗

arXiv2018

The reason for the tilting of domain wall with Dzyaloshinskii-Moriya interaction from a microscopic dynamical perspective

The interfacial Dzyaloshinskii-Moriya interaction (DMI) of a heavy metal (HM)/ferromagnetic (FM) metal heterostructure is vital to the current-induced domain wall motion (CIDWM) at an ultrahigh velocity. However, strong DMI also tilts the moving domain wall (DW) plane, and the mechanism for this tilting is not quite clear. In this work, we have found that this tilting may be understood based on a micromagnetic calculation from a microscopic dynamical perspective. The DMI-induced antisymmetric moment structure at the two boundaries of the track needs to be paid attention. In the early stage of CIDWM induced by spin-orbit torque, this antisymmetry is destroyed. Afterwards, the moments at the two boundaries experience distinct rotation processes with different energy paths towards their final stable antisymmetric moment structure. This results in different initial velocities of the local DW regions at the two boundaries. In mathematics, this distinct DW dynamical progresses at the two boundaries can be approximately revealed by modifying the initial conditions for solving the Thiele equations.

physics.comp-ph↗

arXiv2018

Recycled Least Squares Estimation in Nonlinear Regression

We consider a resampling scheme for parameters estimates in nonlinear regression models. We provide an estimation procedure which recycles, via random weighting, the relevant parameters estimates to construct consistent estimates of the sampling distribution of the various estimates. We establish the asymptotic normality of the resampled estimates and demonstrate the applicability of the recycling approach in a small simulation study and via example.

stat.ME↗

arXiv2018

Broadening Dark Matter Searches at the LHC: Mono-X versus Darkonium Channels

Current searches for dark matter at the LHC focus on mono-X signatures: the production of dark matter in association with a Standard Model (SM) particle. The simplest benchmark introduces a massive spin-1 mediator, the $Z^\prime$ boson, between the dark matter $χ$ and the SM. Limits derived from mono-X channels are most effective when the mediator can decay into two on-shell dark matter particles: $M_{Z'}\gtrsim 2M_χ$. We broaden the experimental reach into the complementary region, where the $Z^\prime$ mediator is much lighter than the dark matter. In this scenario the $Z^\prime$ mediates an effective long-range force between the dark matter, thereby facilitating the formation of darkonium bound states, as is common in many dark sector models. The darkonium becomes active when $M_χ>M_{Z'}/α_{\rm eff}$, where $α_{\rm eff}$ is the effective fine-structure constant in the dark sector. Moreover, the darkonium could decay back into SM quarks, without producing missing transverse momentum in the detector. Considering multijet final states, we reinterpret existing searches to constrain the simple $Z^\prime$ benchmark beyond the region probed by mono-X searches. Assuming a baryonic $Z^\prime$ mediator and a Dirac dark matter, direct detection bounds can be loosened by giving a small Majorana mass to the dark matter. We also consider the interplay between mono-X and darkonium channels at future high energy colliders, which is at the frontier of probing the model parameter space.

hep-ph↗

arXiv2018

Electroweak Baryogenesis From Dark CP Violation

We present a novel mechanism of electroweak baryogenesis where \textit{CP} violation occurs in a dark sector, comprised of standard model gauge singlets, thereby evading the strong electric dipole moment constraints. In this framework, the background of time-like component of a new gauge boson $Z^\prime_μ$, generated at electroweak temperatures, drives the electroweak sphaleron processes to create the required baryon asymmetry. We first discuss the crucial ingredients for this mechanism to work, and then show that all of them can be elegantly embedded in ultraviolet completions with spontaneously broken gauged lepton number. The models under consideration have a rich phenomenology and can be experimentally probed in leptophilic $Z^\prime$ searches, dark matter searches, heavy Majorana neutrino searches, as well as through hunting for new Higgs portal scalars in multi-lepton channels at colliders.

hep-ph↗