arXiv subjects

Wei Xu

Publications and source records attributed to Wei Xu.

At least 631 records · Page 35Linked to original sources

arXiv2017

Emergent eigenstate solution and emergent Gibbs ensemble for expansion dynamics in optical lattices

Within the emergent eigenstate solution to quantum dynamics [Phys. Rev. X 7, 021012 (2017)], one can construct a local operator (an emergent Hamiltonian) of which the time-evolving state is an eigenstate. Here we show that such a solution exists for the expansion dynamics of Tonks-Girardeau gases in optical lattices after turning off power-law (e.g., harmonic or quartic) confining potentials, which are geometric quenches that do not involve the boost operator. For systems that are initially in the ground state and undergo dynamical fermionization during the expansion, we show that they remain in the ground state of the emergent local Hamiltonian at all times. On the other hand, for systems at nonzero initial temperatures, the expansion dynamics can be described constructing a Gibbs ensemble for the emergent local Hamiltonian (an emergent Gibbs ensemble).

cond-mat.quant-gas↗

arXiv2017

Joint Time Allocation and Power Control in Multi-Cell Networks with Load Coupling: Energy Saving and Rate Improvement

In this paper, we consider the problems of minimizing sum power and maximizing sum rate for multi-cell networks with load coupling, where coupling relation occurs among cells due to inter-cell interference. This coupling relation is characterized by the signal-to-interference-and-noise-ratio (SINR) coupling model with cell load vector and cell power vector as the variables. Due to the nonlinear SINR coupling model, the optimization problems for multi-cell networks with load coupling is nonconvex. To solve these nonconvex problems, we first consider the optimization problems for single-cell networks. Through variable transformations, the optimization problems can be equivalently transformed into convex problems. By solving the Karush-Kuhn-Tucker (KKT), the optimal solutions to power minimization and rate maximization problems can be obtained in closed form. Based on the theoretical findings of optimization problems for single-cell networks, we develop a distributed time allocation and power control algorithm with low complexity for sum power minimization in multi-cell networks. This algorithm is proved to be convergent and globally optimal by using the properties of standard interference function. For sum rate optimization in multi-cell networks, we also provide a distributed algorithm which yields suboptimal solution. Besides, the convergence for this distributed algorithm is proved. Numerical results illustrate the theoretical findings, showing the superiority of our solutions compared to the conventional solution of allocating uniform power for users in the same cell.

cs.IT↗

arXiv2017

Spectral and Energy Efficiency of Multi-pair Massive MIMO Relay Network with Hybrid Processing

We consider a multi-pair massive multiple-input multiple-output (MIMO) relay network, where the relay is equipped with a large number, N, of antennas, but driven by a far smaller number, L, of radio frequency (RF) chains. We assume that K pairs of users are scheduled for simultaneous transmission, where K satisfies 2K = L. A hybrid signal processing scheme is presented for both uplink and downlink transmissions of the network. Analytical expressions of both spectral and energy efficiency are derived with respect to the RF chain number under imperfect channel estimation. It is revealed that, under the condition N > 4L^2/pi, the transmit power of each user and the relay can be respectively scaled down by 1=sqrt(N) and 2K=sqrt(N) if pilot power scales with signal power, or they can be respectively scaled down by 1=N and 2K=N if the pilot power is kept fixed, while maintaining an asymptotically unchanged spectral efficiency (SE). While regarding energy efficiency (EE) of the network, the optimal EE is shown to be achieved when Pr = 2KPs, where Pr and Ps respectively refer to the transmit power of the relay and each source terminal. We show that the network EE is a quasi-concave function with respect to the number of RF-chains which, therefore, admits a unique globally optimal choice of the RF-chain number. Numerical simulations are conducted to verify our observations.

cs.IT↗

arXiv2017

On Performance of Quantized Transceiver in Multiuser Massive MIMO Downlinks

Low-resolution digital-to-analog converters (DACs) and analog-to-digital converters (ADCs) are considered to reduce cost and power consumption in multiuser massive multiple-input multiple-output (MIMO). Using the Bussgang theorem, we derive the asymptotic downlink achievable rate w.r.t the resolutions of both DACs and ADCs, i.e., $b_{DA}$ and $b_{AD}$, under the assumption of large antenna number, $N$, and fixed user load ratio, $β$. We characterize the rate loss caused by finite-bit-resolution converters and reveal that the quantization distortion is ignorable at low signal-to-noise ratio (SNR) even with low-resolution converters at both sides. While for maintaining the same rate loss at high SNR, it is discovered that one-more-bit DAC resolution is needed when more users are scheduled with $β$ increased by four times. More specifically for one-bit rate loss requirement, $b_{DA}$ can be set by $\left\lceil b_{AD}+\frac{1}{2}\logβ\right\rceil$ given $b_{AD}$. Similar observations on ADCs are also obtained with numerical verifications.

cs.IT↗

arXiv2017

A Spatial Structural Derivative Model for Ultraslow Diffusion

This study investigates the ultraslow diffusion by a spatial structural derivative, in which the exponential function exp(x)is selected as the structural function to construct the local structural derivative diffusion equation model. The analytical solution of the diffusion equation is a form of Biexponential distribution. Its corresponding mean squared displacement is numerically calculated, and increases more slowly than the logarithmic function of time. The local structural derivative diffusion equation with the structural function exp(x)in space is an alternative physical and mathematical modeling model to characterize a kind of ultraslow diffusion.

cond-mat.stat-mech↗

arXiv2017

Feasibility study on the least square method for fitting non-Gaussian noise data

This study is to investigate the feasibility of least square method in fitting non-Gaussian noise data. We add different levels of the two typical non-Gaussian noises, Lévy and stretched Gaussian noises, to exact value of the selected functions including linear equations, polynomial and exponential equations, and the maximum absolute and the mean square errors are calculated for the different cases. Lévy and stretched Gaussian distributions have many applications in fractional and fractal calculus. It is observed that the non-Gaussian noises are less accurately fitted than the Gaussian noise, but the stretched Gaussian cases appear to perform better than the Lévy noise cases. It is stressed that the least-squares method is inapplicable to the non-Gaussian noise cases when the noise level is larger than 5%.

stat.OT↗

arXiv2017

Listen, Interact and Talk: Learning to Speak via Interaction

One of the long-term goals of artificial intelligence is to build an agent that can communicate intelligently with human in natural language. Most existing work on natural language learning relies heavily on training over a pre-collected dataset with annotated labels, leading to an agent that essentially captures the statistics of the fixed external training data. As the training data is essentially a static snapshot representation of the knowledge from the annotator, the agent trained this way is limited in adaptiveness and generalization of its behavior. Moreover, this is very different from the language learning process of humans, where language is acquired during communication by taking speaking action and learning from the consequences of speaking action in an interactive manner. This paper presents an interactive setting for grounded natural language learning, where an agent learns natural language by interacting with a teacher and learning from feedback, thus learning and improving language skills while taking part in the conversation. To achieve this goal, we propose a model which incorporates both imitation and reinforcement by leveraging jointly sentence and reward feedbacks from the teacher. Experiments are conducted to validate the effectiveness of the proposed approach.

cs.CL↗

arXiv2017

Unsupervised Learning Layers for Video Analysis

This paper presents two unsupervised learning layers (UL layers) for label-free video analysis: one for fully connected layers, and the other for convolutional ones. The proposed UL layers can play two roles: they can be the cost function layer for providing global training signal; meanwhile they can be added to any regular neural network layers for providing local training signals and combined with the training signals backpropagated from upper layers for extracting both slow and fast changing features at layers of different depths. Therefore, the UL layers can be used in either pure unsupervised or semi-supervised settings. Both a closed-form solution and an online learning algorithm for two UL layers are provided. Experiments with unlabeled synthetic and real-world videos demonstrated that the neural networks equipped with UL layers and trained with the proposed online learning algorithm can extract shape and motion information from video sequences of moving objects. The experiments demonstrated the potential applications of UL layers and online learning algorithm to head orientation estimation and moving object localization.

cs.LG↗

arXiv2017

A Deep Compositional Framework for Human-like Language Acquisition in Virtual Environment

We tackle a task where an agent learns to navigate in a 2D maze-like environment called XWORLD. In each session, the agent perceives a sequence of raw-pixel frames, a natural language command issued by a teacher, and a set of rewards. The agent learns the teacher's language from scratch in a grounded and compositional manner, such that after training it is able to correctly execute zero-shot commands: 1) the combination of words in the command never appeared before, and/or 2) the command contains new object concepts that are learned from another task but never learned from navigation. Our deep framework for the agent is trained end to end: it learns simultaneously the visual representations of the environment, the syntax and semantics of the language, and the action module that outputs actions. The zero-shot learning capability of our framework results from its compositionality and modularity with parameter tying. We visualize the intermediate outputs of the framework, demonstrating that the agent truly understands how to solve the problem. We believe that our results provide some preliminary insights on how to train an agent with similar abilities in a 3D environment.

cs.CL↗

arXiv2017

Backward Doubly Stochastic Equations with Jumps and Comparison Theorems

In this work we mainly prove the existence and pathwise uniqueness of solutions to general backward doubly stochastic differential equations with jumps appearing in both forward and backward integral parts. Several comparison theorems under some weak conditions are also given. Finally we apply comparison theorems in proving the existence of solution to some special backward doubly stochastic differential equations with drift coefficient increasing linearly.

math.PR↗

arXiv2017

Neural Machine Translation with Pivot Languages

While recent neural machine translation approaches have delivered state-of-the-art performance for resource-rich language pairs, they suffer from the data scarcity problem for resource-scarce language pairs. Although this problem can be alleviated by exploiting a pivot language to bridge the source and target languages, the source-to-pivot and pivot-to-target translation models are usually independently trained. In this work, we introduce a joint training algorithm for pivot-based neural machine translation. We propose three methods to connect the two models and enable them to interact with each other during training. Experiments on Europarl and WMT corpora show that joint training of source-to-pivot and pivot-to-target models leads to significant improvements over independent training across various languages.

cs.CL↗

arXiv2017

Towards Economic Models for MOOC Pricing Strategy Design

MOOCs have brought unprecedented opportunities of making high-quality courses accessible to everybody. However, from the business point of view, MOOCs are often challenged for lacking of sustainable business models, and academic research for marketing strategies of MOOCs is also a blind spot currently. In this work, we try to formulate the business models and pricing strategies in a structured and scientific way. Based on both theoretical research and real marketing data analysis from a MOOC platform, we present the insights of the pricing strategies for existing MOOC markets. We focus on the pricing strategies for verified certificates in the B2C markets, and also give ideas of modeling the course sub-licensing services in B2B markets.

cs.CY↗

arXiv2016

An Optimization Framework For Online Ride-sharing Markets

Taxi services and product delivery services are instrumental for our modern society. Thanks to the emergence of sharing economy, ride-sharing services such as Uber, Didi, Lyft and Google's Waze Rider are becoming more ubiquitous and grow into an integral part of our everyday lives. However, the efficiency of these services are severely limited by the sub-optimal and imbalanced matching between the supply and demand. We need a generalized framework and corresponding efficient algorithms to address the efficient matching, and hence optimize the performance of these markets. Existing studies for taxi and delivery services are only applicable in scenarios of the one-sided market. In contrast, this work investigates a highly generalized model for the taxi and delivery services in the market economy (abbreviated as"taxi and delivery market") that can be widely used in two-sided markets. Further, we present efficient online and offline algorithms for different applications. We verify our algorithm with theoretical analysis and trace-driven simulations under realistic settings.

cs.SI↗

arXiv2016

Semi-Supervised Learning for Neural Machine Translation

While end-to-end neural machine translation (NMT) has made remarkable progress recently, NMT systems only rely on parallel corpora for parameter estimation. Since parallel corpora are usually limited in quantity, quality, and coverage, especially for low-resource languages, it is appealing to exploit monolingual corpora to improve NMT. We propose a semi-supervised approach for training NMT models on the concatenation of labeled (parallel corpora) and unlabeled (monolingual corpora) data. The central idea is to reconstruct the monolingual corpora using an autoencoder, in which the source-to-target and target-to-source translation models serve as the encoder and decoder, respectively. Our approach can not only exploit the monolingual corpora of the target language, but also of the source language. Experiments on the Chinese-English dataset show that our approach achieves significant improvements over state-of-the-art SMT and NMT systems.

cs.CL↗

arXiv2016

Discovering Conversational Dependencies between Messages in Dialogs

We investigate the task of inferring conversational dependencies between messages in one-on-one online chat, which has become one of the most popular forms of customer service. We propose a novel probabilistic classifier that leverages conversational, lexical and semantic information. The approach is evaluated empirically on a set of customer service chat logs from a Chinese e-commerce website. It outperforms heuristic baselines.

cs.CL↗

arXiv2016

Deep Joint Face Hallucination and Recognition

Deep models have achieved impressive performance for face hallucination tasks. However, we observe that directly feeding the hallucinated facial images into recog- nition models can even degrade the recognition performance despite the much better visualization quality. In this paper, we address this problem by jointly learning a deep model for two tasks, i.e. face hallucination and recognition. In particular, we design an end-to-end deep convolution network with hallucination sub-network cascaded by recognition sub-network. The recognition sub- network are responsible for producing discriminative feature representations using the hallucinated images as inputs generated by hallucination sub-network. During training, we feed LR facial images into the network and optimize the parameters by minimizing two loss items, i.e. 1) face hallucination loss measured by the pixel wise difference between the ground truth HR images and network-generated images; and 2) verification loss which is measured by the classification error and intra-class distance. We extensively evaluate our method on LFW and YTF datasets. The experimental results show that our method can achieve recognition accuracy 97.95% on 4x down-sampled LFW testing set, outperforming the accuracy 96.35% of conventional face recognition model. And on the more challenging YTF dataset, we achieve recognition accuracy 90.65%, a margin over the recognition accuracy 89.45% obtained by conventional face recognition model on the 4x down-sampled version.

cs.CV↗

arXiv2016

Automatically Building Face Datasets of New Domains from Weakly Labeled Data with Pretrained Models

Training data are critical in face recognition systems. However, labeling a large scale face data for a particular domain is very tedious. In this paper, we propose a method to automatically and incrementally construct datasets from massive weakly labeled data of the target domain which are readily available on the Internet under the help of a pretrained face model. More specifically, given a large scale weakly labeled dataset in which each face image is associated with a label, i.e. the name of an identity, we create a graph for each identity with edges linking matched faces verified by the existing model under a tight threshold. Then we use the maximal subgraph as the cleaned data for that identity. With the cleaned dataset, we update the existing face model and use the new model to filter the original dataset to get a larger cleaned dataset. We collect a large weakly labeled dataset containing 530,560 Asian face images of 7,962 identities from the Internet, which will be published for the study of face recognition. By running the filtering process, we obtain a cleaned datasets (99.7+% purity) of size 223,767 (recall 70.9%). On our testing dataset of Asian faces, the model trained by the cleaned dataset achieves recognition rate 93.1%, which obviously outperforms the model trained by the public dataset CASIA whose recognition rate is 85.9%.

cs.CV↗

arXiv2016

A critical point in Sr2-xIrO4 and less distorted IrO6 octahedra induced by deep Sr-vacancies

For the Sr2-xIrO4 system, recent work pointed out the abnormal electronic and magnetic properties around the critical point. Here, to understand Sr-vacancy effect on spin-orbit coupled Mott insulator Sr2IrO4, the electronic structure and local structure distortion for Sr2-xIrO4 system have been investigated by x-ray absorption spectroscopy (XAS). By comparing the intensity of white-line features at the IrL2,3-edge x-ray absorption near-edge spectroscopy (XANES), we observe a sudden rise of the branching ratio in the vicinity of the critical point.Further analysis on theIrL3 -edge extended x-ray absorption fine structure (EXAFS) and calculated data demonstrated that the abrupt enhancement of the branching ratio is intimately related to the less distorted IrO6 octahedra induced by deep Sr-vacancies, which in turn alters the relatively strength of the spin-orbit coupling (SOC) and crystal electric field (CEF)that would dictate the abnormal behavior of electronic and magnetic structure near the critical point.

cond-mat.str-el↗