arXiv subjects

Shuai Li

Publications and source records attributed to Shuai Li.

At least 415 records · Page 23Linked to original sources

arXiv2017

Full-duplex Amplify-and-Forward Relaying: Power and Location Optimization

In this paper, we consider a full-duplex (FD) amplify-and-forward (AF) relay system and optimize its power allocation and relay location to minimize the system symbol error rate (SER). We first derive the asymptotic expressions of the outage probability and SER performance by taking into account the residual self interference (RSI) in FD systems. We then formulate the optimization problem based on the minimal SER criterion. Analytical and numerical results show that optimized relay location and power allocation can greatly improve system SER performance, and the performance floor caused by the RSI can be significantly reduced via optimizing relay location or power allocation.

cs.IT↗

arXiv2017

Deep LSTM for Large Vocabulary Continuous Speech Recognition

Recurrent neural networks (RNNs), especially long short-term memory (LSTM) RNNs, are effective network for sequential task like speech recognition. Deeper LSTM models perform well on large vocabulary continuous speech recognition, because of their impressive learning ability. However, it is more difficult to train a deeper network. We introduce a training framework with layer-wise training and exponential moving average methods for deeper LSTM models. It is a competitive framework that LSTM models of more than 7 layers are successfully trained on Shenma voice search data in Mandarin and they outperform the deep LSTM models trained by conventional approach. Moreover, in order for online streaming speech recognition applications, the shallow model with low real time factor is distilled from the very deep model. The recognition accuracy have little loss in the distillation process. Therefore, the model trained with the proposed training framework reduces relative 14\% character error rate, compared to original model which has the similar real-time capability. Furthermore, the novel transfer learning strategy with segmental Minimum Bayes-Risk is also introduced in the framework. The strategy makes it possible that training with only a small part of dataset could outperform full dataset training from the beginning.

cs.CL↗

arXiv2017

Fano-shaped impurity spectral density, electric-field-induced in-gap state and local magnetic moment of an adatom on trilayer graphene

Recently, the existence of local magnetic moment in a hydrogen adatom on graphene has been confirmed experimentally [González-Herrero et al., Science, 2016, 352, 437]. Inspired by this breakthrough, we theoretically investigate the top-site adatom on trilayer graphene (TLG) by solving the Anderson impurity model via self-consistent mean field method. The influence of the stacking order, the adsorption site and external electric field are carefully considered. We find that, due to its unique electronic structure, the situation of the TLG is drastically different from that of the monolayer graphene. Firstly, the adatom on rhombohedral stacked TLG (r-TLG) can have a Fano-shaped impurity spectral density, instead of the normal Lorentzian-like one, when the impurity level is around the Fermi level. Secondly, the impurity level of the adatom on r-TLG can be tuned into an in-gap state by an external electric field, which strongly depends on the direction of the applied electric field and can significantly affect the local magnetic moment formation. Finally, we systematically calculate the impurity magnetic phase diagrams, considering various stacking orders, adsorption sites, doping and electric field. We show that, because of the in-gap state, the impurity magnetic phase of r-TLG will obviously depend on the direction of the applied electric field as well. All our theoretical results can be readily tested in experiment, and may give a comprehensive understanding about the local magnetic moment of adatom on TLG.

cond-mat.mes-hall↗

arXiv2017

On Context-Dependent Clustering of Bandits

We investigate a novel cluster-of-bandit algorithm CAB for collaborative recommendation tasks that implements the underlying feedback sharing mechanism by estimating the neighborhood of users in a context-dependent manner. CAB makes sharp departures from the state of the art by incorporating collaborative effects into inference as well as learning processes in a manner that seamlessly interleaving explore-exploit tradeoffs and collaborative steps. We prove regret bounds under various assumptions on the data, which exhibit a crisp dependence on the expected number of clusters over the users, a natural measure of the statistical difficulty of the learning task. Experiments on production and real-world datasets show that CAB offers significantly increased prediction performance against a representative pool of state-of-the-art methods.

cs.LG↗

arXiv2017

Context-Aware Bandits

We propose an efficient Context-Aware clustering of Bandits (CAB) algorithm, which can capture collaborative effects. CAB can be easily deployed in a real-world recommendation system, where multi-armed bandits have been shown to perform well in particular with respect to the cold-start problem. CAB utilizes a context-aware clustering augmented by exploration-exploitation strategies. CAB dynamically clusters the users based on the content universe under consideration. We give a theoretical analysis in the standard stochastic multi-armed bandits setting. We show the efficiency of our approach on production and real-world datasets, demonstrate the scalability, and, more importantly, the significant increased prediction performance against several state-of-the-art methods.

cs.LG↗

arXiv2017

X-duplex Relaying: Adaptive Antenna Configuration

In this letter, we propose a joint transmission mode and transmit/receive (Tx/Rx) antenna configuration scheme referred to as X-duplex in the relay network with one source, one amplify-and-forward (AF) relay and one destination. The relay is equipped with two antennas, each of which is capable of reception and transmission. In the proposed scheme, the relay adaptively selects its Tx and Rx antenna, operating in either full-duplex (FD) or half-duplex (HD) mode. The proposed scheme is based on minimizing the symbol error rate (SER) of the relay system. The asymptotic expressions of the cumulative distribution function (CDF) for the end-to-end signal to interference plus noise ratio (SINR), average SER and diversity order are derived and validated by simulations. Results show that the X-duplex scheme achieves additional spatial diversity, significantly reduces the performance floor at high SNR and improves the system performance.

cs.IT↗

arXiv2017

On the Performance of X-Duplex Relaying

In this paper, we study a X-duplex relay system with one source, one amplify-and-forward (AF) relay and one destination, where the relay is equipped with a shared antenna and two radio frequency (RF) chains used for transmission or reception. X-duplex relay can adaptively configure the connection between its RF chains and antenna to operate in either HD or FD mode, according to the instantaneous channel conditions. We first derive the distribution of the signal to interference plus noise ratio (SINR), based on which we then analyze the outage probability, average symbol error rate (SER), and average sum rate. We also investigate the X-duplex relay with power allocation and derive the lower bound and upper bound of the corresponding outage probability. Both analytical and simulated results show that the X-duplex relay achieves a better performance over pure FD and HD schemes in terms of SER, outage probability and average sum rate, and the performance floor caused by the residual self interference can be eliminated using flexible RF chain configurations.

cs.NI↗

arXiv2017

Automatic Discoveries of Physical and Semantic Concepts via Association Priors of Neuron Groups

The recent successful deep neural networks are largely trained in a supervised manner. It {\it associates} complex patterns of input samples with neurons in the last layer, which form representations of {\it concepts}. In spite of their successes, the properties of complex patterns associated a learned concept remain elusive. In this work, by analyzing how neurons are associated with concepts in supervised networks, we hypothesize that with proper priors to regulate learning, neural networks can automatically associate neurons in the intermediate layers with concepts that are aligned with real world concepts, when trained only with labels that associate concepts with top level neurons, which is a plausible way for unsupervised learning. We develop a prior to verify the hypothesis and experimentally find the proposed prior help neural networks automatically learn both basic physical concepts at the lower layers, e.g., rotation of filters, and highly semantic concepts at the higher layers, e.g., fine-grained categories of an entry-level category.

cs.LG↗

arXiv2017

Akid: A Library for Neural Network Research and Production from a Dataism Approach

Neural networks are a revolutionary but immature technique that is fast evolving and heavily relies on data. To benefit from the newest development and newly available data, we want the gap between research and production as small as possibly. On the other hand, differing from traditional machine learning models, neural network is not just yet another statistic model, but a model for the natural processing engine --- the brain. In this work, we describe a neural network library named {\texttt akid}. It provides higher level of abstraction for entities (abstracted as blocks) in nature upon the abstraction done on signals (abstracted as tensors) by Tensorflow, characterizing the dataism observation that all entities in nature processes input and emit out in some ways. It includes a full stack of software that provides abstraction to let researchers focus on research instead of implementation, while at the same time the developed program can also be put into production seamlessly in a distributed environment, and be production ready. At the top application stack, it provides out-of-box tools for neural network applications. Lower down, akid provides a programming paradigm that lets user easily build customized models. The distributed computing stack handles the concurrency and communication, thus letting models be trained or deployed to a single GPU, multiple GPUs, or a distributed environment without affecting how a model is specified in the programming paradigm stack. Lastly, the distributed deployment stack handles how the distributed computing is deployed, thus decoupling the research prototype environment with the actual production environment, and is able to dynamically allocate computing resources, so development (Devs) and operations (Ops) could be separated. Please refer to http://akid.readthedocs.io/en/latest/ for documentation.

cs.LG↗

arXiv2016

From Simplicity to Complexity Based on Consensus: A Case Study

Distributed consensus has been intensively studied in recent years as a means to mitigate state differences among dynamic nodes on a graph. It has been successfully employed in various applications, e.g., formation control of multi-robots, load balancing, clock synchronization. However, almost all existing applications cast an impression of consensus as a simple process to iteratively reach agreement, without any clue on possibility to generate advanced complexity, say shortest path planning, which has been proved to be NP-hard. Counter-intuitively, we show for the first time that the complexity of shortest path planning can emerge from a perturbed version of min-consensus protocol, which as a case study may shed lights to researchers in the field of distributed control to re-think the nature of complexity and the distance between control and intelligence. Besides, we rigorously prove the convergence of graph dynamics and its equivalence to shortest path solutions. An illustrative simulation on a small scale graph is provided to show the convergence of the biased min-consensus dynamics to shortest path solution over the graph. To demonstrate the scalability to large scale problems, a graph with 43826 nodes, which corresponds to a map of a maze in 2D, is considered in the simulation study. Apart from possible applications in robot path planning, the result is further extended to robot complete coverage, showing its potential in real practice such as cleaning robots.

eess.SY↗

arXiv2016

Desiging Artificial Lieb Lattice on Metal Surface

Recently, several experiments have illustrated that metal surface electrons can be manipulated to form a two dimensional (2D) lattice by depositing a designer molecule lattice on metal surface. This offers a promising new technique to construct artificial 2D electron lattices. Here we theoretically propose a molecule lattice pattern to realize an artificial Lieb lattice on metal surface, which shows a flat electronic band due to the lattice geometry. We show that the localization of electrons in the flat band may be understood from the viewpoint of electron interference, which may be probed by measuring the local density of states with the scanning tunnelling microscopy. Our proposal may be readily implemented in experiment and may offer an ideal solid state platform to investigate the novel flat band physics of the Lieb lattice.

cond-mat.mes-hall↗

arXiv2016

Online Optimization Methods for the Quantification Problem

The estimation of class prevalence, i.e., the fraction of a population that belongs to a certain class, is a very useful tool in data analytics and learning, and finds applications in many domains such as sentiment analysis, epidemiology, etc. For example, in sentiment analysis, the objective is often not to estimate whether a specific text conveys a positive or a negative sentiment, but rather estimate the overall distribution of positive and negative sentiments during an event window. A popular way of performing the above task, often dubbed quantification, is to use supervised learning to train a prevalence estimator from labeled data. Contemporary literature cites several performance measures used to measure the success of such prevalence estimators. In this paper we propose the first online stochastic algorithms for directly optimizing these quantification-specific performance measures. We also provide algorithms that optimize hybrid performance measures that seek to balance quantification and classification performance. Our algorithms present a significant advancement in the theory of multivariate optimization and we show, by a rigorous theoretical analysis, that they exhibit optimal convergence. We also report extensive experiments on benchmark and real data sets which demonstrate that our methods significantly outperform existing optimization techniques used for these performance measures.

stat.ML↗

arXiv2016

Tapered Simplified Modal Method for Analysis of Non-rectangular Gratings

The Simplified Modal Method (SMM) provides a quick and intuitive way to analyze the performance of gratings of rectangular shapes. For non-rectangular shapes, a version of SMM has been developed, but it applies only to the Littrow-mounting incidence case and it neglects reflection. Here, we use the theory of mode-coupling in a tapered waveguide to improve SMM so that it applies to non-rectangular gratings at arbitrary angles of incidence. Moreover, this new 'Tapered Simplified Modal Method' (TSMM) allows us to properly account for reflected light. We present here the analytical development of the theory and numerical simulations, demonstrating the validity of the method.

physics.optics↗

arXiv2016

Distributed Clustering of Linear Bandits in Peer to Peer Networks

We provide two distributed confidence ball algorithms for solving linear bandit problems in peer to peer networks with limited communication capabilities. For the first, we assume that all the peers are solving the same linear bandit problem, and prove that our algorithm achieves the optimal asymptotic regret rate of any centralised algorithm that can instantly communicate information between the peers. For the second, we assume that there are clusters of peers solving the same bandit problem within each cluster, and we prove that our algorithm discovers these clusters, while achieving the optimal asymptotic regret rate within each one. Through experiments on several real-world datasets, we demonstrate the performance of proposed algorithms compared to the state-of-the-art.

cs.LG↗

arXiv2016

Collaborative Filtering Bandits

Classical collaborative filtering, and content-based filtering methods try to learn a static recommendation model given training data. These approaches are far from ideal in highly dynamic recommendation domains such as news recommendation and computational advertisement, where the set of items and users is very fluid. In this work, we investigate an adaptive clustering technique for content recommendation based on exploration-exploitation strategies in contextual multi-armed bandit settings. Our algorithm takes into account the collaborative effects that arise due to the interaction of the users with the items, by dynamically grouping users based on the items under consideration and, at the same time, grouping items based on the similarity of the clusterings induced over the users. The resulting algorithm thus takes advantage of preference patterns in the data in a way akin to collaborative filtering methods. We provide an empirical analysis on medium-size real-world datasets, showing scalability and increased prediction performance (as measured by click-through rate) over state-of-the-art methods for clustering bandits. We also provide a regret analysis within a standard linear stochastic noise setting.

cs.LG↗

arXiv2016

Graph Clustering Bandits for Recommendation

We investigate an efficient context-dependent clustering technique for recommender systems based on exploration-exploitation strategies through multi-armed bandits over multiple users. Our algorithm dynamically groups users based on their observed behavioral similarity during a sequence of logged activities. In doing so, the algorithm reacts to the currently served user by shaping clusters around him/her but, at the same time, it explores the generation of clusters over users which are not currently engaged. We motivate the effectiveness of this clustering policy, and provide an extensive empirical analysis on real-world datasets, showing scalability and improved prediction performance over state-of-the-art methods for sequential clustering of users in multi-armed bandit scenarios.

stat.ML↗

arXiv2016

The Effective Model of the Molecule Graphene System and One Application Beyond Graphene

Recently, a new kind of two dimensional (2D) artificial electron lattice, i.e. molecule graphene, has drawn a lots of interest, where the metal surface electrons are transformed into a honeycomb lattice via absorbing a molecule lattice on metal surface. In this work, we would like to point out that the technique used to build molecule graphene actually gives a promising way to explore the interesting physics of other novel 2D lattice beyond graphene. The key issue is that this system is an antidot system, where the absorbed molecule normally gives a repulsive potential. So, we need to establish a map between the molecule arrangement and the desired surface electron lattice. To give a concrete example, we first analyse the effective models of molecule graphene, and estimate the model parameters for the Cu/CO system through numerical fitting the exerimental data. Then, we design a molecule lattice, and show that, with this kind of lattice, the surface electrons can be transformed into a Kagome like lattice. Using the estimated parameters of Cu/CO system, we calculate the corresponding energy bands and LDOS, which can be readily tested in experiment. We hope that our work can stimulate further theoretical and experimental interest in this novel artificial 2D electron lattice.

cond-mat.mes-hall↗

arXiv2016

Influence of surface conductivity on the apparent zeta potential of calcite

Zeta potential is a physicochemical parameter of particular importance in describing the surface electrical properties of charged porous media. However, the zeta potential of calcite is still poorly known because of the difficulty to interpret streaming potential experiments. The Helmholtz-Smoluchowski (HS) equation is widely used to estimate the apparent zeta potential from these experiments. However, this equation neglects the influence of surface conductivity on streaming potential. We present streaming potential and electrical conductivity measurements on a calcite powder in contact with an aqueous NaCl electrolyte. Our streaming potential model corrects the apparent zeta potential of calcite by accounting for the influence of surface conductivity and flow regime. We show that the HS equation seriously underestimates the zeta potential of calcite, particularly when the electrolyte is diluted (ionic strength < 0.01 M) because of calcite surface conductivity. The basic Stern model successfully predicted the corrected zeta potential by assuming that the zeta potential is located at the outer Helmholtz plane, i.e. without considering a stagnant diffuse layer at the calcite-water interface. The surface conductivity of calcite crystals was inferred from electrical conductivity measurements and computed using our basic Stern model. Surface conductivity was also successfully predicted by our surface complexation model.

physics.geo-ph↗