arXiv subjects

Jian Yang

Publications and source records attributed to Jian Yang.

At least 757 records · Page 42Linked to original sources

arXiv2017

Particle-Hole Symmetry and the Fractional Quantum Hall States at 5/2 Filling Factor

We propose a derivative operator formed as a function of derivatives of the electron coordinates. When the derivative operator is applied to the Laughlin wave function, two new wave functions in the lowest Landau level at filling factor 1/2 are generated. For systems of 4, 6, and 8 electrons in spherical geometry, it is shown that the first wave function has nearly unity overlap with the particle-hole conjugate of the Moore-Read Pfaffian wave function, therefore together with the Moore-Read Pfaffian state forms a particle-hole conjugate pair. The second wave function has essentially perfect particle-hole symmetry itself, with a positive parity when the number of electron pairs N/2 is an even integer and and a negative parity when N/2 is an odd integer. An equivalent form suggests the first wave function forms a f-wave pairing of composite fermions, and the second wave function forms a p-wave pairing. The corresponding Non-Abelian statistics quasiparticle wave functions are also proposed.

cond-mat.str-el↗

arXiv2016

LightRNN: Memory and Computation-Efficient Recurrent Neural Networks

Recurrent neural networks (RNNs) have achieved state-of-the-art performances in many natural language processing tasks, such as language modeling and machine translation. However, when the vocabulary is large, the RNN model will become very big (e.g., possibly beyond the memory capacity of a GPU device) and its training will become very inefficient. In this work, we propose a novel technique to tackle this challenge. The key idea is to use 2-Component (2C) shared embedding for word representations. We allocate every word in the vocabulary into a table, each row of which is associated with a vector, and each column associated with another vector. Depending on its position in the table, a word is jointly represented by two components: a row vector and a column vector. Since the words in the same row share the row vector and the words in the same column share the column vector, we only need $2 \sqrt{|V|}$ vectors to represent a vocabulary of $|V|$ unique words, which are far less than the $|V|$ vectors required by existing approaches. Based on the 2-Component shared embedding, we design a new RNN algorithm and evaluate it using the language modeling task on several benchmark datasets. The results show that our algorithm significantly reduces the model size and speeds up the training process, without sacrifice of accuracy (it achieves similar, if not better, perplexity as compared to state-of-the-art language models). Remarkably, on the One-Billion-Word benchmark Dataset, our algorithm achieves comparable perplexity to previous language models, whilst reducing the model size by a factor of 40-100, and speeding up the training process by a factor of 2. We name our proposed algorithm \emph{LightRNN} to reflect its very small model size and very high training speed.

cs.CL↗

arXiv2016

Entropy-based link prediction in weighted networks

Information entropy has been proved to be an effective tool to quantify the structural importance of complex networks. In the previous work (Xu et al, 2016 \cite{xu2016}), we measure the contribution of a path in link prediction with information entropy. In this paper, we further quantify the contribution of a path with both path entropy and path weight, and propose a weighted prediction index based on the contributions of paths, namely Weighted Path Entropy (WPE), to improve the prediction accuracy in weighted networks. Empirical experiments on six weighted real-world networks show that WPE achieves higher prediction accuracy than three typical weighted indices.

cs.SI↗

arXiv2016

A Link between Sequential Semi-anonymous Nonatomic Games and their Large Finite Counterparts

We show that equilibria of a sequential semi-anonymous nonatomic game (SSNG) can be adopted by players in corresponding large but finite dynamic games to achieve near-equilibrium payoffs. Such equilibria in the form of random state-to-action rules are parsimonious in form and easy to execute, as they are both oblivious of past history and blind to other players' present states. Our transient results can be extended to a stationary case, where the finite counterparts are special discounted stochastic games. The kind of equilibria we adopt for SSNG are similar to distributional equilibria that are well understood in literature, and they themselves are shown to exist.

econ.GN↗

arXiv2016

Traffic dynamics on dynamical networks: The connection between network lifetime and traffic congestion

For many power-limited networks, such as wireless sensor networks and mobile ad hoc networks, maximizing the network lifetime is the first concern in the related designing and maintaining activities. We study the network lifetime from the perspective of network science. In our dynamic network, nodes are assigned a fixed amount of energy initially and consume the energy in the delivery of packets. We divided the network traffic flow into four states: no, slow, fast, and absolute congestion states. We derive the network lifetime by considering the state of the traffic flow. We find that the network lifetime is generally opposite to traffic congestion in that the more congested traffic, the less network lifetime. We also find the impacts of factors such as packet generation rate, communication radius, node moving speed, etc., on network lifetime and traffic congestion.

cs.NI↗

arXiv2016

A survey of sparse representation: algorithms and applications

Sparse representation has attracted much attention from researchers in fields of signal processing, image processing, computer vision and pattern recognition. Sparse representation also has a good reputation in both theoretical research and practical applications. Many different algorithms have been proposed for sparse representation. The main purpose of this article is to provide a comprehensive study and an updated review on sparse representation and to supply a guidance for researchers. The taxonomy of sparse representation methods can be studied from various viewpoints. For example, in terms of different norm minimizations used in sparsity constraints, the methods can be roughly categorized into five groups: sparse representation with $l_0$-norm minimization, sparse representation with $l_p$-norm (0$<$p$<$1) minimization, sparse representation with $l_1$-norm minimization and sparse representation with $l_{2,1}$-norm minimization. In this paper, a comprehensive overview of sparse representation is provided. The available sparse representation algorithms can also be empirically categorized into four groups: greedy strategy approximation, constrained optimization, proximity algorithm-based optimization, and homotopy algorithm-based sparse representation. The rationales of different algorithms in each category are analyzed and a wide range of sparse representation applications are summarized, which could sufficiently reveal the potential nature of the sparse representation theory. Specifically, an experimentally comparative study of these sparse representation algorithms was presented. The Matlab code used in this paper can be available at: http://www.yongxu.org/lunwen.html.

cs.CV↗

arXiv2015

Link prediction based on path entropy

Information theory has been taken as a prospective tool for quantifying the complexity of complex networks. In this paper, we first study the information entropy or uncertainty of a path using the information theory. Then we apply the path entropy to the link prediction problem in real-world networks. Specifically, we propose a new similarity index, namely Path Entropy (PE) index, which considers the information entropies of shortest paths between node pairs with penalization to long paths. Empirical experiments demonstrate that PE index outperforms the mainstream link predictors.

physics.soc-ph↗

arXiv2015

Inventory Control Involving Unknown Demand of Discrete Nonperishable Items - Analysis of a Newsvendor-based Policy

Inventory control with unknown demand distribution is considered, with emphasis placed on the case involving discrete nonperishable items. We focus on an adaptive policy which in every period uses, as much as possible, the optimal newsvendor ordering quantity for the empirical distribution learned up to that period. The policy is assessed using the regret criterion, which measures the price paid for ambiguity on demand distribution over $T$ periods. When there are guarantees on the latter's separation from the critical newsvendor parameter $β=b/(h+b)$, a constant upper bound on regret can be found. Without any prior information on the demand distribution, we show that the regret does not grow faster than the rate $T^{1/2+ε}$ for any $ε>0$. In view of a known lower bound, this is almost the best one could hope for. Simulation studies involving this along with other policies are also conducted.

stat.ML↗

arXiv2015

Information transport in multiplex networks

In this paper, we study information transport in multiplex networks comprised of two coupled subnetworks. The upper subnetwork, called the logical layer, employs the shortest paths protocol to determine the logical paths for packets transmission, while the lower subnetwork acts as the physical layer, in which packets are delivered by the biased random walk mechanism characterized with a parameter $α$. Through simulation, we obtain the optimal $α$ corresponding to the maximum network lifetime and the maximum number of the arrival packets. Assortative coupling is better than the random coupling and the disassortative coupling, since it achieves much better transmission performances. Generally, the more homogeneous the lower subnetwork, the better the transmission performances are, which is opposite for the upper subnetwork. Finally, we propose an attack centrality for nodes based on the topological information of both subnetworks, and further investigate the transmission performances under targeted attacks. Our work helps to understand the spreading and robustness issues of multiplex networks and provides some clues about the designing of more efficient and robust routing architectures in communication systems.

physics.soc-ph↗

arXiv2015

Stochastic Behavior of the Nonnegative Least Mean Fourth Algorithm for Stationary Gaussian Inputs and Slow Learning

Some system identification problems impose nonnegativity constraints on the parameters to estimate due to inherent physical characteristics of the unknown system. The nonnegative least-mean-square (NNLMS) algorithm and its variants allow to address this problem in an online manner. A nonnegative least mean fourth (NNLMF) algorithm has been recently proposed to improve the performance of these algorithms in cases where the measurement noise is not Gaussian. This paper provides a first theoretical analysis of the stochastic behavior of the NNLMF algorithm for stationary Gaussian inputs and slow learning. Simulation results illustrate the accuracy of the proposed analysis.

math.NA↗

arXiv2015

Traffic-driven SIR epidemic model on networks

We propose a novel SIR epidemic model which is driven by the transmission of infection packets in networks. Specifically, infected nodes generate and deliver infection packets causing the spread of the epidemic, while recovered nodes block the delivery of infection packets, and this inhibits the epidemic spreading. The efficient routing protocol governed by a control parameter $α$ is used in the packet transmission. We obtain the maximum instantaneous population of infected nodes, the maximum population of ever infected nodes, as well as the corresponding optimal $α$ through simulation. We find that generally more balanced load distribution leads to more intense and wide spread of an epidemic in networks. Increasing either average node degree or homogeneity of degree distribution will facilitate epidemic spreading. When packet generation rate $ρ$ is small, increasing $ρ$ favors epidemic spreading. However, when $ρ$ is large enough, traffic congestion appears which inhibits epidemic spreading.

physics.soc-ph↗

arXiv2015

Entangling Color-Different Photons via Time-Resolved Measurement and Active Feed-Forward

Entangling independent photons is not only of fundamental interest but also of crucial importance for quantum information science. Two-photon interference is a major method to entangle independent identical photons. If two photons are color-different, perfect two-photon coalescence cannot happen anymore, which makes the entangling of color-different photons difficult to realize. In this letter by exploring and developing time-resolved measurement and active feed-forward, we have entangled two independent photons of different colors for the first time. We find that entanglement with a varying form can be identified for different two-photon temporal modes through time-resolved measurement. By using active feed-forward we are able to convert the varying entanglement into uniform. Adopting these measures, we have successfully entangled two photons with a frequency separation of 16 times larger than their linewidths. In addition to its fundamental interest, our work also provides an approach to solve the frequency mismatch problem for future quantum networks.

quant-ph↗

arXiv2015

Sparse Deep Stacking Network for Image Classification

Sparse coding can learn good robust representation to noise and model more higher-order representation for image classification. However, the inference algorithm is computationally expensive even though the supervised signals are used to learn compact and discriminative dictionaries in sparse coding techniques. Luckily, a simplified neural network module (SNNM) has been proposed to directly learn the discriminative dictionaries for avoiding the expensive inference. But the SNNM module ignores the sparse representations. Therefore, we propose a sparse SNNM module by adding the mixed-norm regularization (l1/l2 norm). The sparse SNNM modules are further stacked to build a sparse deep stacking network (S-DSN). In the experiments, we evaluate S-DSN with four databases, including Extended YaleB, AR, 15 scene and Caltech101. Experimental results show that our model outperforms related classification methods with only a linear classifier. It is worth noting that we reach 98.8% recognition accuracy on 15 scene.

cs.CV↗

arXiv2014

Feature Selection in Conditional Random Fields for Map Matching of GPS Trajectories

Map matching of the GPS trajectory serves the purpose of recovering the original route on a road network from a sequence of noisy GPS observations. It is a fundamental technique to many Location Based Services. However, map matching of a low sampling rate on urban road network is still a challenging task. In this paper, the characteristics of Conditional Random Fields with regard to inducing many contextual features and feature selection are explored for the map matching of the GPS trajectories at a low sampling rate. Experiments on a taxi trajectory dataset show that our method may achieve competitive results along with the success of reducing model complexity for computation-limited applications.

stat.ML↗

arXiv2014

Feature Engineering for Map Matching of Low-Sampling-Rate GPS Trajectories in Road Network

Map matching of GPS trajectories from a sequence of noisy observations serves the purpose of recovering the original routes in a road network. In this work in progress, we attempt to share our experience of feature construction in a spatial database by reporting our ongoing experiment of feature extrac-tion in Conditional Random Fields (CRFs) for map matching. Our preliminary results are obtained from real-world taxi GPS trajectories.

stat.ML↗

arXiv2014

Epidemic spreading driven by biased random walks

Random walk is one of the basic mechanisms found in many network applications. We study the epidemic spreading dynamics driven by biased random walks on complex networks. In our epidemic model, each time infected nodes constantly spread some infected packets by biased random walks to their neighbor nodes causing the infection of the susceptible nodes that receive the packets. An infected node get recovered from infection with a fixed probability. Simulation and analytical results on model and real-world networks show that the epidemic spreading becomes intense and wide with the increase of delivery capacity of infected nodes, average node degree, homogeneity of node degree distribution. Furthermore, there are corresponding optimal parameters such that the infected nodes have instantaneously the largest population, and the epidemic spreading process covers the largest part of a network.

physics.soc-ph↗

arXiv2014

Unsupervised Pretraining Encourages Moderate-Sparseness

It is well known that direct training of deep neural networks will generally lead to poor results. A major progress in recent years is the invention of various pretraining methods to initialize network parameters and it was shown that such methods lead to good prediction performance. However, the reason for the success of pretraining has not been fully understood, although it was argued that regularization and better optimization play certain roles. This paper provides another explanation for the effectiveness of pretraining, where we show pretraining leads to a sparseness of hidden unit activation in the resulting neural networks. The main reason is that the pretraining models can be interpreted as an adaptive sparse coding. Compared to deep neural network with sigmoid function, our experimental results on MNIST and Birdsong further support this sparseness observation.

cs.LG↗

arXiv2014

Nuclear Norm based Matrix Regression with Applications to Face Recognition with Occlusion and Illumination Changes

Recently regression analysis becomes a popular tool for face recognition. The existing regression methods all use the one-dimensional pixel-based error model, which characterizes the representation error pixel by pixel individually and thus neglects the whole structure of the error image. We observe that occlusion and illumination changes generally lead to a low-rank error image. To make use of this low-rank structural information, this paper presents a two-dimensional image matrix based error model, i.e. matrix regression, for face representation and classification. Our model uses the minimal nuclear norm of representation error image as a criterion, and the alternating direction method of multipliers method to calculate the regression coefficients. Compared with the current regression methods, the proposed Nuclear Norm based Matrix Regression (NMR) model is more robust for alleviating the effect of illumination, and more intuitive and powerful for removing the structural noise caused by occlusion. We experiment using four popular face image databases, the Extended Yale B database, the AR database, the Multi-PIE and the FRGC database. Experimental results demonstrate the performance advantage of NMR over the state-of-the-art regression based face recognition methods.

cs.CV↗