arXiv subjects

Jie Wang

Publications and source records attributed to Jie Wang.

At least 595 records · Page 33Linked to original sources

arXiv2019

Edge Induced Topological Phase Transition of the Quantum Hall state at Half Filling

We show that in quantum Hall systems at half-filling, edge potentials alone can drive transitions between the Pfaffian and anti-Pfaffian topological phases. We conjecture this is true in realistic systems even in the presence of weak bulk interactions that break the particle-hole symmetry. The strong effects of edge potentials could be understood from different topological shifts of competing phases, manifested on the disk geometry as the variation of orbital numbers at fixed number of particles. In particular, we show analytically particle-hole conjugation of Hamiltonians on the disk is equivalent to the tuning of edge potentials, which allows us to explicitly demonstrate the phase transition numerically. The importance of edge potentials in various experimental contexts, including the recently discovered particle-hole symmetric phase is also discussed.

cond-mat.str-el↗

arXiv2019

Semi-Supervised Self-Taught Deep Learning for Finger Bones Segmentation

Segmentation stands at the forefront of many high-level vision tasks. In this study, we focus on segmenting finger bones within a newly introduced semi-supervised self-taught deep learning framework which consists of a student network and a stand-alone teacher module. The whole system is boosted in a life-long learning manner wherein each step the teacher module provides a refinement for the student network to learn with newly unlabeled data. Experimental results demonstrate the superiority of the proposed method over conventional supervised deep learning methods.

cs.CV↗

arXiv2019

Multi-Agent Deep Reinforcement Learning for Large-scale Traffic Signal Control

Reinforcement learning (RL) is a promising data-driven approach for adaptive traffic signal control (ATSC) in complex urban traffic networks, and deep neural networks further enhance its learning power. However, centralized RL is infeasible for large-scale ATSC due to the extremely high dimension of the joint action space. Multi-agent RL (MARL) overcomes the scalability issue by distributing the global control to each local RL agent, but it introduces new challenges: now the environment becomes partially observable from the viewpoint of each local agent due to limited communication among agents. Most existing studies in MARL focus on designing efficient communication and coordination among traditional Q-learning agents. This paper presents, for the first time, a fully scalable and decentralized MARL algorithm for the state-of-the-art deep RL agent: advantage actor critic (A2C), within the context of ATSC. In particular, two methods are proposed to stabilize the learning procedure, by improving the observability and reducing the learning difficulty of each local agent. The proposed multi-agent A2C is compared against independent A2C and independent Q-learning algorithms, in both a large synthetic traffic grid and a large real-world traffic network of Monaco city, under simulated peak-hour traffic dynamics. Results demonstrate its optimality, robustness, and sample efficiency over other state-of-the-art decentralized MARL algorithms.

cs.LG↗

arXiv2019

Mexican Hat Wavelet Kernel ELM for Multiclass Classification

Kernel extreme learning machine (KELM) is a novel feedforward neural network, which is widely used in classification problems. To some extent, it solves the existing problems of the invalid nodes and the large computational complexity in ELM. However, the traditional KELM classifier usually has a low test accuracy when it faces multiclass classification problems. In order to solve the above problem, a new classifier, Mexican Hat wavelet KELM classifier, is proposed in this paper. The proposed classifier successfully improves the training accuracy and reduces the training time in the multiclass classification problems. Moreover, the validity of the Mexican Hat wavelet as a kernel function of ELM is rigorously proved. Experimental results on different data sets show that the performance of the proposed classifier is significantly superior to the compared classifiers.

cs.LG↗

arXiv2019

A New Sparse SOS Decomposition Algorithm Based on Term Sparsity

A new sparse SOS decomposition algorithm is proposed based on a new sparsity pattern, called cross sparsity patterns. The new sparsity pattern focuses on the sparsity of terms and thus is different from the well-known correlative sparsity pattern which focuses on the sparsity of variables though the sparse SOS decomposition algorithms based on these two sparsity patterns both take use of chordal extensions/chordal decompositions. Moreover, it is proved that the SOS decomposition obtained by the new sparsity pattern is always a refinement of the block-diagonalization obtained by the sign-symmetry method. %Because the new sparsity pattern covers more sparse polynomials than correlative sparsity pattern, Various experiments show that the new algorithm dramatically saves the computational cost compared to existing tools and can handle some really huge polynomials.

math.OC↗

arXiv2018

Rapid Cycle-Accurate Simulator for High-Level Synthesis

A large semantic gap between the high-level synthesis (HLS) design and the low-level (on-board or RTL) simulation environment often creates a barrier for those who are not FPGA experts. Moreover, such low-level simulation takes a long time to complete. Software-based HLS simulators can help bridge this gap and accelerate the simulation process; however, we found that the current FPGA HLS commercial software simulators sometimes produce incorrect results. In order to solve this correctness issue while maintaining the high speed of a software-based simulator, this paper proposes a new HLS simulation flow named FLASH. The main idea behind the proposed flow is to extract the scheduling information from the HLS tool and automatically construct an equivalent cycle-accurate simulation model while preserving C semantics. Experimental results show that FLASH runs three orders of magnitude faster than the RTL simulation.

cs.AR↗

arXiv2018

The abundance of satellite galaxies in the inner region of $Λ$CDM Milky Way sized haloes

The concordance $Λ$CDM cosmology predicts tens of satellite galaxies distributed in the inner region ($<40\ \lkpc$) of the Milky Way (MW), yet at most $12$ were discovered at present day, including 3 discovered very recently by Dark Energy Survey (DES) and 5 from other surveys (e.g. PanSTARRS, MagLiteS). We use $5$ ultra-high resolution simulations of MW sized dark matter haloes from the $\aqua$ project, combined with $\galf$ semi-analytical galaxy formation model, to investigate properties of the model satellite galaxy population inside $40\ \lkpc$ of MW sized haloes. On average, in each halo this model predicts about $20$ inner satellite galaxies, among them $5$ are comparable to the classic satellites in the luminosity, these are in stark contrast to the corresponding numbers in observations. We further investigate the survivability of these model inner satellites in the presence of a central stellar disk with a set of ideal simulations. These are done by re-evolving a quarter (30) of the whole $\aqua$ inner satellite galaxies (121) by including a static disk potential in addition to the MW halo. Our finding is that the additional disk completely disrupt 40 percents of these satellites and results in $14$ satellite galaxies within the $40\ \lkpc$ of each $\aqua$ at the end, in reasonably well agreement with observations.

astro-ph.GA↗

arXiv2018

Randomized Consensus based Distributed Kalman Filtering over Wireless Sensor Networks

This paper is concerned with developing a novel distributed Kalman filtering algorithm over wireless sensor networks based on randomized consensus strategy. Compared with the centralized algorithm, distributed filtering techniques require less computation per sensor and lead to more robust estimation since they simply use the information from the neighboring nodes in the network. However, poor local sensor estimation caused by limited observability and network topology changes which interfere the global consensus are challenging issues. Motivated by this observation, we propose a novel randomized gossip-based distributed Kalman filtering algorithm. Information exchange and computation in the proposed algorithm can be carried out in an arbitrarily connected network of nodes. In addition, the computational burden can be distributed for a sensor which communicates with a stochastically selected neighbor at each clock step under schemes of gossip algorithm. In this case, the error covariance matrix changes stochastically at every clock step, thus the convergence is considered in a probabilistic sense. We provide the mean square convergence analysis of the proposed algorithm. Under a sufficient condition, we show that the proposed algorithm is quite appealing as it achieves better mean square error performance theoretically than the noncooperative decentralized Kalman filtering algorithm. Besides, considering the limited computation, communication, and energy resources in the wireless sensor networks, we propose an optimization problem which minimizes the average expected state estimation error based on the proposed algorithm. To solve the proposed problem efficiently, we transform it into a convex optimization problem. And a sub-optimal solution is attained. Examples and simulations are provided to illustrate the theoretical results.

eess.SY↗

arXiv2018

On Supports of Sums of Nonnegative Circuit Polynomials

In this paper, we prove that every SONC polynomial decomposes into a sum of nonnegative circuit polynomials with the same support, which reveals the advantage of SONC decompositions for certifying nonnegativity of sparse polynomials compared with the classical SOS decompositions. By virtue of this fact, we can decide $f\in$ SONC through relative entropy programming more efficiently.

math.CO↗

arXiv2018

Examining Scientific Writing Styles from the Perspective of Linguistic Complexity

Publishing articles in high-impact English journals is difficult for scholars around the world, especially for non-native English-speaking scholars (NNESs), most of whom struggle with proficiency in English. In order to uncover the differences in English scientific writing between native English-speaking scholars (NESs) and NNESs, we collected a large-scale data set containing more than 150,000 full-text articles published in PLoS between 2006 and 2015. We divided these articles into three groups according to the ethnic backgrounds of the first and corresponding authors, obtained by Ethnea, and examined the scientific writing styles in English from a two-fold perspective of linguistic complexity: (1) syntactic complexity, including measurements of sentence length and sentence complexity; and (2) lexical complexity, including measurements of lexical diversity, lexical density, and lexical sophistication. The observations suggest marginal differences between groups in syntactical and lexical complexity.

cs.CL↗

arXiv2018

Semantic WordRank: Generating Finer Single-Document Summarizations

We present Semantic WordRank (SWR), an unsupervised method for generating an extractive summary of a single document. Built on a weighted word graph with semantic and co-occurrence edges, SWR scores sentences using an article-structure-biased PageRank algorithm with a Softplus function adjustment, and promotes topic diversity using spectral subtopic clustering under the Word-Movers-Distance metric. We evaluate SWR on the DUC-02 and SummBank datasets and show that SWR produces better summaries than the state-of-the-art algorithms over DUC-02 under common ROUGE measures. We then show that, under the same measures over SummBank, SWR outperforms each of the three human annotators (aka. judges) and compares favorably with the combined performance of all judges.

cs.CL↗

arXiv2018

Effective Predictions of Gaokao Admission Scores for College Applications in Mainland China

Gaokao is the annual academic qualification examination for college admissions in mainland China. Organized by each provincial-level administrative region (PAR), Gaokao takes place at the same time nationwide in early June. To enroll in a university in September, students must take Gaokao and submit common applications for admission to their home PAR Gaokao office in July, listing a small and fixed number of universities and majors they intend to attend and study. About 9.5 million high- school seniors participate in Gaokao every year, and the Gaokao scores are good for just one year. A student has a strong chance to be accepted if their Gaokao score is better than the admission scores of the universities they selected in their applications. However, the admission scores of universities are unknown at the time when filling out applications, which to be determined dynamically during the admission process and may fluctuate from year to year. To increase their chances of acceptance to a best- suited university, students need to predict admission score of each university they are interested in. Early prediction methods are empirical without the backing of in-depth data studies. We fill this void by presenting well-tested mathematical models based on the ranking of Gaokao scores in a PAR. We show that our methods significantly outperform the methods commonly used by teachers and experts, and can predict admission scores with an accuracy of 91% within a 7-point margin in an exam of a 750-point grading scale.

cs.CY↗

arXiv2018

Precise Condition Synthesis for Program Repair

Due to the difficulty of repairing defect, many research efforts have been devoted into automatic defect repair. Given a buggy program that fails some test cases, a typical automatic repair technique tries to modify the program to make all tests pass. However, since the test suites in real world projects are usually insufficient, aiming at passing the test suites often leads to incorrect patches. In this paper we aim to produce precise patches, that is, any patch we produce has a relatively high probability to be correct. More concretely, we focus on condition synthesis, which was shown to be able to repair more than half of the defects in existing approaches. Our key insight is threefold. First, it is important to know what variables in a local context should be used in an "if" condition, and we propose a sorting method based on the dependency relations between variables. Second, we observe that the API document can be used to guide the repair process, and propose document analysis technique to further filter the variables. Third, it is important to know what predicates should be performed on the set of variables, and we propose to mine a set of frequently used predicates in similar contexts from existing projects. We develop a novel program repair system, ACS, that could generate precise conditions at faulty locations. Furthermore, given the generated conditions are very precise, we can perform a repair operation that is previously deemed to be too overfitting: directly returning the test oracle to repair the defect. Using our approach, we successfully repaired 17 defects on four projects of Defects4J, which is the largest number of fully automatically repaired defects reported on the dataset so far. More importantly, the precision of our approach in the evaluation is 73.9%, which is significantly higher than previous approaches, which are usually less than 40%.

cs.SE↗

arXiv2018

Infection Analysis on Irregular Networks through Graph Signal Processing

In a networked system, functionality can be seriously endangered when nodes are infected, due to internal random failures or a contagious virus that develops into an epidemic. Given a snapshot of the network representing the nodes' states (infected or healthy), infection analysis refers to distinguishing an epidemic from random failures and gathering information for effective countermeasure design. This analysis is challenging due to irregular network structure, heterogeneous epidemic spreading, and noisy observations. This paper treats a network snapshot as a graph signal, and develops effective approaches for infection analysis based on graph signal processing. For the macro (network-level) analysis aiming to distinguish an epidemic from random failures, 1) multiple detection metrics are defined based on the graph Fourier transform (GFT) and neighborhood characteristics of the graph signal; 2) a new class of graph wavelets, distance-based graph wavelets (DBGWs), are developed; and 3) a machine learning-based framework is designed employing either the GFT spectrum or the graph wavelet coefficients as features for infection analysis. DBGWs also enable the micro (node-level) infection analysis, through which the performance of epidemic countermeasures can be improved. Extensive simulations are conducted to demonstrate the effectiveness of all the proposed algorithms in various network settings.

cs.SI↗

arXiv2018

The effect of air exposure on SEY and surface composition of laser treated copper applied in accelerators

In this paper, the effect of air exposure and acetone cleaning on the secondary electron yield (SEY) of laser treated copper were investigated and discussed. The SEY of laser treated copper samples after acetone cleaning and air exposure were measured. After 10 to 21 months air exposure, the maximum SEY increased by about 11%~20%. Furthermore, changes of carbon, copper, and oxygen chemistry states from sample surface after acetone and air exposure were studied by means of XPS. The XPS results show that the concentration of Cu decreased by about 61%~70% after acetone cleaning and decreased by a further 36%~70% after 10 to 21 months air exposure. Moreover, the XPS measurement reflects the significant increase of the total percentages of C and O after acetone cleaning, and further increase after 10 to 21 months air exposure. The increase of SEY is explained in terms of the introduction of impurities, such as oxide, carbide etc.

physics.acc-ph↗

arXiv2018

Nonlinear Shape Regression For Filtering Segmentation Results From Calcium Imaging

A shape filter is presented to repair segmentation results obtained in calcium imaging of neurons in vivo. This post-segmentation algorithm can automatically smooth the shapes obtained from a preliminary segmentation, while precluding the cases where two neurons are counted as one combined component. The shape filter is realized using a square-root velocity to project the shapes on a shape manifold in which distances between shapes are based on elastic changes. Two data-driven weighting methods are proposed to achieve a trade-off between shape smoothness and consistency with the data. Intuitive comparisons of proposed methods via projection onto Cartesian maps demonstrate the smoothing ability of the shape filter. Quantitative measures also prove the superiority of our methods over models that do not employ any weighting criterion.

eess.IV↗

arXiv2018

Density-dependent clustering: I. Pulling back the curtains on motions of the BAO peak

The most common statistic used to analyze large-scale structure surveys is the correlation function, or power spectrum. Here, we show how `slicing' the correlation function on local density brings sensitivity to interesting non-Gaussian features in the large-scale structure, such as the expansion or contraction of baryon acoustic oscillations (BAO) according to the local density. The sliced correlation function measures the large-scale flows that smear out the BAO, instead of just correcting them as reconstruction algorithms do. Thus, we expect the sliced correlation function to be useful in constraining the growth factor, and modified gravity theories that involve the local density. Out of the studied cases, we find that the run of the BAO peak location with density is best revealed when slicing on a $\sim 40$ Mpc/$h$ filtered density. But slicing on a $\sim100$ Mpc/$h$ filtered density may be most useful in distinguishing between underdense and overdense regions, whose BAO peaks are separated by a substantial $\sim 5$ Mpc/$h$ at $z=0$. We also introduce `curtain plots' showing how local densities drive particle motions toward or away from each other over the course of an $N$-body simulation.

astro-ph.CO↗

arXiv2018

Resolution of the apparent discrepancy between the number of massive subhaloes in Abell 2744 and ΛCDM

Schwinn et al. (2017) have recently compared the abundance and distribution of massive substructures identified in a gravitational lensing analysis of Abell 2744 by Jauzac et al. (2016) and N-body simulation and found no cluster in ΛCDM simulation that is similar to Abell 2744. Schwinn et al.(2017) identified the measured projected aperture masses with the actual masses associated with subhaloes in the MXXL N-body simulation. We have used the high resolution Phoenix cluster simulations to show that such an identification is incorrect: the aperture mass is dominated by mass in the body of the cluster that happens to be projected along the line-of-sight to the subhalo. This enhancement varies from factors of a few to factors of more than 100, particularly for subhaloes projected near the centre of the cluster. We calculate aperture masses for subhaloes in our simulation and compare them to the measurements for Abell 2744. We find that the data for Abell 2744 are in excellent agreement with the matched predictions from ΛCDM. We provide further predictions for aperture mass functions of subhaloes in idealized surveys with varying mass detection thresholds.

astro-ph.GA↗