arXiv subjects

Chen Liang

Publications and source records attributed to Chen Liang.

191 records · Page 11Linked to original sources

arXiv2018

Liquid structure changes of Bi under different melt states

This paper is focused on the relationship between solidification and melt sates. An abnormal change which takes place at 744 C from 84 to 120 min indicated a liquid-liquid structure transition in pure Bi melt. Based on liquid-liquid structure transition, experiments of pure Bi were carried out. Cooling from different melt states, the solidification behaviors and structures changed a lot, such as undercooling, solidification time, grain sizes and the number of twins and pedestal sits. Especially, as a semi-metal, a liquid-liquid structure transition greatly increases the metallicity of Bi, and during the solidification Bi turned from smooth to coarse solid/ liquid interface.

cond-mat.mtrl-sci↗

arXiv2018

Failure Localization in Power Systems via Tree Partitions

Cascading failures in power systems propagate non-locally, making the control and mitigation of outages extremely hard. In this work, we use the emerging concept of the tree partition of transmission networks to provide an analytical characterization of line failure localizability in transmission systems. Our results rigorously establish the well perceived intuition in power community that failures cannot cross bridges, and reveal a finer-grained concept that encodes more precise information on failure propagations within tree-partition regions. Specifically, when a non-bridge line is tripped, the impact of this failure only propagates within well-defined components, which we refer to as cells, of the tree partition defined by the bridges. In contrast, when a bridge line is tripped, the impact of this failure propagates globally across the network, affecting the power flow on all remaining transmission lines. This characterization suggests that it is possible to improve the system robustness by temporarily switching off certain transmission lines, so as to create more, smaller components in the tree partition; thus spatially localizing line failures and making the grid less vulnerable to large-scale outages. We illustrate this approach using the IEEE 118-bus test system and demonstrate that switching off a negligible portion of transmission lines allows the impact of line failures to be significantly more localized without substantial changes in line congestion.

eess.SY↗

arXiv2018

A Fully Convolutional Tri-branch Network (FCTN) for Domain Adaptation

A domain adaptation method for urban scene segmentation is proposed in this work. We develop a fully convolutional tri-branch network, where two branches assign pseudo labels to images in the unlabeled target domain while the third branch is trained with supervision based on images in the pseudo-labeled target domain. The re-labeling and re-training processes alternate. With this design, the tri-branch network learns target-specific discriminative representations progressively and, as a result, the cross-domain capability of the segmenter improves. We evaluate the proposed network on large-scale domain adaptation experiments using both synthetic (GTA) and real (Cityscapes) images. It is shown that our solution achieves the state-of-the-art performance and it outperforms previous methods by a significant margin.

cs.CV↗

arXiv2018

Active Learning of Strict Partial Orders: A Case Study on Concept Prerequisite Relations

Strict partial order is a mathematical structure commonly seen in relational data. One obstacle to extracting such type of relations at scale is the lack of large-scale labels for building effective data-driven solutions. We develop an active learning framework for mining such relations subject to a strict order. Our approach incorporates relational reasoning not only in finding new unlabeled pairs whose labels can be deduced from an existing label set, but also in devising new query strategies that consider the relational structure of labels. Our experiments on concept prerequisite relations show our proposed framework can substantially improve the classification performance with the same query budget compared to other baseline approaches.

cs.LG↗

arXiv2017

High energy properties of the flat spectrum radio quasar 4C 50.11

We investigate the $γ$-ray and X-ray properties of the flat spectrum radio quasar (FSRQ) \4c50 at redshift $z= 1.517$. The {\it Fermi} Large Area Telescope (LAT) data indicate that this source was in an active state since 2013 July. During the active period, the $γ$-ray flux increased by more than a factor of three and two distinct flares were detected with the variability timescale as short as several hours. The $γ$-ray spectra can be well fitted by a log-parabola. From the fitting, we find a correlation between the peak energy and spectral curvature for the $γ$-ray spectra, which is the first time seen in $γ$-ray emission from a blazar. The {\it Swift} X-ray Telescope (XRT) data show that the source was variable at X-ray energies, but no evidence shows flux or spectral changes related to the $γ$-ray activity. The broad-band X-ray spectrum obtained with {\it Swift} XRT and {\it NuSTAR} is well described by a broken PL model, with an extremely hard spectrum ($Γ_{1} \sim 0.1$) below the break energy, $E_{\rm break} \sim 2.1 {\rm keV}$, and $Γ_{2} \sim 1.5$ above the break energy. The spectral steepening below $\sim 3$ keV is likely due to the low energy cut-off in the energy distribution of the photon-emitting electron population. Both the $γ$-ray and X-ray emission appear harder when brighter. The broad-band spectral energy distribution (SED) is constructed for the source, and we provide a model fit to the SED. Our modeling suggests that the emission region should be outside the broad line region, and the properties of the region indicate a jet with 42\% of the Eddington power during the active state.

astro-ph.HE↗

arXiv2017

Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision

Harnessing the statistical power of neural networks to perform language understanding and symbolic reasoning is difficult, when it requires executing efficient discrete operations against a large knowledge-base. In this work, we introduce a Neural Symbolic Machine, which contains (a) a neural "programmer", i.e., a sequence-to-sequence model that maps language utterances to programs and utilizes a key-variable memory to handle compositionality (b) a symbolic "computer", i.e., a Lisp interpreter that performs program execution, and helps find good programs by pruning the search space. We apply REINFORCE to directly optimize the task reward of this structured prediction problem. To train with weak supervision and improve the stability of REINFORCE, we augment it with an iterative maximum-likelihood training process. NSM outperforms the state-of-the-art on the WebQuestionsSP dataset when trained from question-answer pairs only, without requiring any feature engineering or domain-specific knowledge.

cs.CL↗

arXiv2016

Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervision (Short Version)

Extending the success of deep neural networks to natural language understanding and symbolic reasoning requires complex operations and external memory. Recent neural program induction approaches have attempted to address this problem, but are typically limited to differentiable memory, and consequently cannot scale beyond small synthetic tasks. In this work, we propose the Manager-Programmer-Computer framework, which integrates neural networks with non-differentiable memory to support abstract, scalable and precise operations through a friendly neural computer interface. Specifically, we introduce a Neural Symbolic Machine, which contains a sequence-to-sequence neural "programmer", and a non-differentiable "computer" that is a Lisp interpreter with code assist. To successfully apply REINFORCE for training, we augment it with approximate gold programs found by an iterative maximum likelihood training process. NSM is able to learn a semantic parser from weak supervision over a large knowledge base. It achieves new state-of-the-art performance on WebQuestionsSP, a challenging semantic parsing dataset, with weak supervision. Compared to previous approaches, NSM is end-to-end, therefore does not rely on feature engineering or domain specific knowledge.

cs.CL↗

arXiv2016

Definition Modeling: Learning to define word embeddings in natural language

Distributed representations of words have been shown to capture lexical semantics, as demonstrated by their effectiveness in word similarity and analogical relation tasks. But, these tasks only evaluate lexical semantics indirectly. In this paper, we study whether it is possible to utilize distributed representations to generate dictionary definitions of words, as a more direct and transparent representation of the embeddings' semantics. We introduce definition modeling, the task of generating a definition for a given word and its embedding. We present several definition model architectures based on recurrent neural networks, and experiment with the models over multiple data sets. Our results show that a model that controls dependencies between the word being defined and the definition words performs significantly better, and that a character-level convolution layer designed to leverage morphology can complement word-level embeddings. Finally, an error analysis suggests that the errors made by a definition model may provide insight into the shortcomings of word embeddings.

cs.CL↗

arXiv2016

Finding Needles in the Haystack: Harnessing Syslogs for Data Center Management

Network device syslogs are ubiquitous and abundant in modern data centers with most large data centers producing millions of messages per day. Yet, the operational information reflected in syslogs and their implications on diagnosis or management tasks are poorly understood. Prevalent approaches to understanding syslogs focus on simple correlation and abnormality detection and are often limited to detection providing little insight towards diagnosis and resolution. Towards improving data center operations, we propose and implement Log-Prophet, a system that applies a toolbox of statistical techniques and domain-specific models to mine detailed diagnoses. Log-Prophet infers causal relationships between syslog lines and constructs succinct but valuable problem graphs, summarizing root causes and their locality, including cascading problems. We validate Log-Prophet using problem tickets and through operator interviews. To demonstrate the strength of Log-Prophet, we perform an initial longitudinal study of a large online service provider's data center. Our study demonstrates that Log-Prophet significantly reduces the number of alerts while highlighting interesting operational issues.

cs.NI↗

arXiv2012

The Connection between Radio and Gamma Ray Emission in Fermi/LAT Blazars

We collect the 2LAC and MOJAVE quasi-simultaneous data to investigate the radio-gamma connection of blazars. The cross sample contains 166 sources. The statistic analysis based on this sample confirms positive correlations between these two bands, but the correlations become weaker as the gamma-ray energy increases. The statistic results between various parameters show negative correlations of gamma-ray photon spectral index with gamma-ray loudness for both FSRQs and BL Lacertae objects, positive correlations of gamma-ray variability index with the gamma-ray loudness for FSRQs, a negative correlation of the gamma-ray variability index with the gamma-ray photon spectral index for FSRQs, and negative correlations of gamma-ray photon spectral index with gamma-ray luminosity for FSRQs. These results suggest that the gamma-ray variability may be due to changes inside the gamma-ray emission region like the injected power, rather than changes in the photon density of the external radiation fields, and the variability amplitude tends to be larger as the gamma-rays are closer to the high energy peak of spectral energy distribution. No correlation of variability index found for BL Lacertae objects implies that variability behavior may differ below and above the peak energy.

astro-ph.HE↗

arXiv2010

Is Low-Frequency-Peaked BL Lac Object OJ 287 a TeV Emitter?

It is well known that there are only two low-frequency-peaked BL Lac objects (LBLs: BL Lacertae and S5 0716+714) and one flat spectrum radio quasar (FSRQ: 3C 279) among more than 30 active galactic nuclei (AGNs) with detected TeV emissions. We study the spectral energy distribution (SED) of a famous LBL OJ 287, whose light curve has a 12-y period. Using a homogeneous one-zone synchrotron + synchrotron-self Compton model, we model the quasi-simultaneous broad-band SED of OJ 287. With some reasonable assumptions, we extrapolate the model to the high state of OJ 287 and predict its γ-ray emissions. Taking into account the absorption of γ-ray by the extragalactic background light (EBL), we find that the TeV emission of OJ 287 in high state is slightly higher than the sensitivity of H.E.S.S. The study on SEDs of OJ 287 has implications to unveil the origin of jet activity during its 12-y period and the properties of EBL.

astro-ph.CO↗