arXiv subjects

Jian Tang

Publications and source records attributed to Jian Tang.

At least 361 records · Page 20Linked to original sources

arXiv2018

On the Universal Approximation Property and Equivalence of Stochastic Computing-based Neural Networks and Binary Neural Networks

Large-scale deep neural networks are both memory intensive and computation-intensive, thereby posing stringent requirements on the computing platforms. Hardware accelerations of deep neural networks have been extensively investigated in both industry and academia. Specific forms of binary neural networks (BNNs) and stochastic computing based neural networks (SCNNs) are particularly appealing to hardware implementations since they can be implemented almost entirely with binary operations. Despite the obvious advantages in hardware implementation, these approximate computing techniques are questioned by researchers in terms of accuracy and universal applicability. Also it is important to understand the relative pros and cons of SCNNs and BNNs in theory and in actual hardware implementations. In order to address these concerns, in this paper we prove that the "ideal" SCNNs and BNNs satisfy the universal approximation property with probability 1 (due to the stochastic behavior). The proof is conducted by first proving the property for SCNNs from the strong law of large numbers, and then using SCNNs as a "bridge" to prove for BNNs. Based on the universal approximation property, we further prove that SCNNs and BNNs exhibit the same energy complexity. In other words, they have the same asymptotic energy consumption with the growing of network size. We also provide a detailed analysis of the pros and cons of SCNNs and BNNs for hardware implementations and conclude that SCNNs are more suitable for hardware.

cs.LG↗

arXiv2018

Model-Free Control for Distributed Stream Data Processing using Deep Reinforcement Learning

In this paper, we focus on general-purpose Distributed Stream Data Processing Systems (DSDPSs), which deal with processing of unbounded streams of continuous data at scale distributedly in real or near-real time. A fundamental problem in a DSDPS is the scheduling problem with the objective of minimizing average end-to-end tuple processing time. A widely-used solution is to distribute workload evenly over machines in the cluster in a round-robin manner, which is obviously not efficient due to lack of consideration for communication delay. Model-based approaches do not work well either due to the high complexity of the system environment. We aim to develop a novel model-free approach that can learn to well control a DSDPS from its experience rather than accurate and mathematically solvable system models, just as a human learns a skill (such as cooking, driving, swimming, etc). Specifically, we, for the first time, propose to leverage emerging Deep Reinforcement Learning (DRL) for enabling model-free control in DSDPSs; and present design, implementation and evaluation of a novel and highly effective DRL-based control framework, which minimizes average end-to-end tuple processing time by jointly learning the system environment via collecting very limited runtime statistics data and making decisions under the guidance of powerful Deep Neural Networks. To validate and evaluate the proposed framework, we implemented it based on a widely-used DSDPS, Apache Storm, and tested it with three representative applications. Extensive experimental results show 1) Compared to Storm's default scheduler and the state-of-the-art model-based method, the proposed framework reduces average tuple processing by 33.5% and 14.0% respectively on average. 2) The proposed framework can quickly reach a good scheduling solution during online learning, which justifies its practicability for online control in DSDPSs.

cs.DC↗

arXiv2018

Study of Non-Standard Charged-Current Interactions at the MOMENT experiment

MuOn-decay MEdium baseline NeuTrino beam experiment (MOMENT) is a next-generation accelerator neutrino experiment looking for more physics study. We try to simulate neutrino oscillations confronting with Charged-Current\&Non-Standard neutrino Interactions(CC-NSIs) at MOMENT. These NSIs could alter neutrino production and detection processes and get involved in neutrino oscillation channels. We separate a perturbative discussion of oscillation channels at near and far detectors, and analyze parameter correlations with the impact of CC-NSIs. Taking $δ_{cp}$ and $θ_{23}$ as an example, we find that CC-NSIs can induce bias in precision measurements of standard oscillation parameters. In addition, a combination of near and far detectors using Gd-doped water cherenkov technology at MOMENT is able to provide good constraints of CC-NSIs happening at the neutrino production and detection processes.

hep-ph↗

arXiv2018

Towards Ultra-High Performance and Energy Efficiency of Deep Learning Systems: An Algorithm-Hardware Co-Optimization Framework

Hardware accelerations of deep learning systems have been extensively investigated in industry and academia. The aim of this paper is to achieve ultra-high energy efficiency and performance for hardware implementations of deep neural networks (DNNs). An algorithm-hardware co-optimization framework is developed, which is applicable to different DNN types, sizes, and application scenarios. The algorithm part adopts the general block-circulant matrices to achieve a fine-grained tradeoff between accuracy and compression ratio. It applies to both fully-connected and convolutional layers and contains a mathematically rigorous proof of the effectiveness of the method. The proposed algorithm reduces computational complexity per layer from O($n^2$) to O($n\log n$) and storage complexity from O($n^2$) to O($n$), both for training and inference. The hardware part consists of highly efficient Field Programmable Gate Array (FPGA)-based implementations using effective reconfiguration, batch processing, deep pipelining, resource re-using, and hierarchical control. Experimental results demonstrate that the proposed framework achieves at least 152X speedup and 71X energy efficiency gain compared with IBM TrueNorth processor under the same test accuracy. It achieves at least 31X energy efficiency gain compared with the reference FPGA-based work.

cs.LG↗

arXiv2018

Deep Reinforcement Learning for Dynamic Treatment Regimes on Medical Registry Data

This paper presents the first deep reinforcement learning (DRL) framework to estimate the optimal Dynamic Treatment Regimes from observational medical data. This framework is more flexible and adaptive for high dimensional action and state spaces than existing reinforcement learning methods to model real-life complexity in heterogeneous disease progression and treatment choices, with the goal of providing doctor and patients the data-driven personalized decision recommendations. The proposed DRL framework comprises (i) a supervised learning step to predict the most possible expert actions, and (ii) a deep reinforcement learning step to estimate the long-term value function of Dynamic Treatment Regimes. Both steps depend on deep neural networks. As a key motivational example, we have implemented the proposed framework on a data set from the Center for International Bone Marrow Transplant Research (CIBMTR) registry database, focusing on the sequence of prevention and treatments for acute and chronic graft versus host disease after transplantation. In the experimental results, we have demonstrated promising accuracy in predicting human experts' decisions, as well as the high expected reward function in the DRL-based dynamic treatment regimes.

cs.AI↗

arXiv2018

Experience-driven Networking: A Deep Reinforcement Learning based Approach

Modern communication networks have become very complicated and highly dynamic, which makes them hard to model, predict and control. In this paper, we develop a novel experience-driven approach that can learn to well control a communication network from its own experience rather than an accurate mathematical model, just as a human learns a new skill (such as driving, swimming, etc). Specifically, we, for the first time, propose to leverage emerging Deep Reinforcement Learning (DRL) for enabling model-free control in communication networks; and present a novel and highly effective DRL-based control framework, DRL-TE, for a fundamental networking problem: Traffic Engineering (TE). The proposed framework maximizes a widely-used utility function by jointly learning network environment and its dynamics, and making decisions under the guidance of powerful Deep Neural Networks (DNNs). We propose two new techniques, TE-aware exploration and actor-critic-based prioritized experience replay, to optimize the general DRL framework particularly for TE. To validate and evaluate the proposed framework, we implemented it in ns-3, and tested it comprehensively with both representative and randomly generated network topologies. Extensive packet-level simulation results show that 1) compared to several widely-used baseline methods, DRL-TE significantly reduces end-to-end delay and consistently improves the network utility, while offering better or comparable throughput; 2) DRL-TE is robust to network changes; and 3) DRL-TE consistently outperforms a state-ofthe-art DRL method (for continuous control), Deep Deterministic Policy Gradient (DDPG), which, however, does not offer satisfying performance.

cs.NI↗

arXiv2018

A Systematic Analysis for State-of-the-Art 3D Lung Nodule Proposals Generation

Lung nodule proposals generation is the primary step of lung nodule detection and has received much attention in recent years . In this paper, we first construct a model of 3-dimension Convolutional Neural Network (3D CNN) to generate lung nodule proposals, which can achieve the state-of-the-art performance. Then, we analyze a series of key problems concerning the training performance and efficiency. Firstly, we train the 3D CNN model with data in different resolutions and find out that models trained by high resolution input data achieve better lung nodule proposals generation performances especially for nodules in too small sizes, while consumes much more memory at the same time. Then, we analyze the memory consumptions on different platforms and the experimental results indicate that CPU architecture can provide us with larger memory and enables us to explore more possibilities of 3D applications. We implement the 3D CNN model on CPU platform and propose an Intel Extended-Caffe framework which supports many highly-efficient 3D computations, which is opened source at https://github.com/extendedcaffe/extended-caffe.

cs.CV↗

arXiv2017

Adversarial Network Embedding

Learning low-dimensional representations of networks has proved effective in a variety of tasks such as node classification, link prediction and network visualization. Existing methods can effectively encode different structural properties into the representations, such as neighborhood connectivity patterns, global structural role similarities and other high-order proximities. However, except for objectives to capture network structural properties, most of them suffer from lack of additional constraints for enhancing the robustness of representations. In this paper, we aim to exploit the strengths of generative adversarial networks in capturing latent features, and investigate its contribution in learning stable and robust graph representations. Specifically, we propose an Adversarial Network Embedding (ANE) framework, which leverages the adversarial learning principle to regularize the representation learning. It consists of two components, i.e., a structure preserving component and an adversarial learning component. The former component aims to capture network structural properties, while the latter contributes to learning robust representations by matching the posterior distribution of the latent representations to given priors. As shown by the empirical results, our method is competitive with or superior to state-of-the-art approaches on benchmark network embedding tasks.

cs.LG↗

arXiv2017

SpecWatch: A Framework for Adversarial Spectrum Monitoring with Unknown Statistics

In cognitive radio networks (CRNs), dynamic spectrum access has been proposed to improve the spectrum utilization, but it also generates spectrum misuse problems. One common solution to these problems is to deploy monitors to detect misbehaviors on certain channel. However, in multi-channel CRNs, it is very costly to deploy monitors on every channel. With a limited number of monitors, we have to decide which channels to monitor. In addition, we need to determine how long to monitor each channel and in which order to monitor, because switching channels incurs costs. Moreover, the information about the misuse behavior is not available a priori. To answer those questions, we model the spectrum monitoring problem as an adversarial multi-armed bandit problem with switching costs (MAB-SC), propose an effective framework, and design two online algorithms, SpecWatch-II and SpecWatch-III, based on the same framework. To evaluate the algorithms, we use weak regret, i.e., the performance difference between the solution of our algorithm and optimal (fixed) solution in hindsight, as the metric. We prove that the expected weak regret of SpecWatch-II is O(T^{2/3}), where T is the time horizon. Whereas, the actual weak regret of SpecWatch-III is O(T^{2/3}) with probability 1 - δ, for any δ in (0, 1). Both algorithms guarantee the upper bounds matching the lower bound of the general adversarial MAB- SC problem. Therefore, they are all asymptotically optimal.

cs.IT↗

arXiv2017

Theoretical Properties for Neural Networks with Weight Matrices of Low Displacement Rank

Recently low displacement rank (LDR) matrices, or so-called structured matrices, have been proposed to compress large-scale neural networks. Empirical results have shown that neural networks with weight matrices of LDR matrices, referred as LDR neural networks, can achieve significant reduction in space and computational complexity while retaining high accuracy. We formally study LDR matrices in deep learning. First, we prove the universal approximation property of LDR neural networks with a mild condition on the displacement operators. We then show that the error bounds of LDR neural networks are as efficient as general neural networks with both single-layer and multiple-layer structure. Finally, we propose back-propagation based training algorithm for general LDR neural networks.

cs.LG↗

arXiv2017

An Attention-based Collaboration Framework for Multi-View Network Representation Learning

Learning distributed node representations in networks has been attracting increasing attention recently due to its effectiveness in a variety of applications. Existing approaches usually study networks with a single type of proximity between nodes, which defines a single view of a network. However, in reality there usually exists multiple types of proximities between nodes, yielding networks with multiple views. This paper studies learning node representations for networks with multiple views, which aims to infer robust node representations across different views. We propose a multi-view representation learning approach, which promotes the collaboration of different views and lets them vote for the robust representations. During the voting process, an attention mechanism is introduced, which enables each node to focus on the most informative views. Experimental results on real-world networks show that the proposed approach outperforms existing state-of-the-art approaches for network representation learning with a single view and other competitive approaches with multiple views.

cs.SI↗

arXiv2017

End-to-end Learning for Short Text Expansion

Effectively making sense of short texts is a critical task for many real world applications such as search engines, social media services, and recommender systems. The task is particularly challenging as a short text contains very sparse information, often too sparse for a machine learning algorithm to pick up useful signals. A common practice for analyzing short text is to first expand it with external information, which is usually harvested from a large collection of longer texts. In literature, short text expansion has been done with all kinds of heuristics. We propose an end-to-end solution that automatically learns how to expand short text to optimize a given learning task. A novel deep memory network is proposed to automatically find relevant information from a collection of longer documents and reformulate the short text through a gating mechanism. Using short text classification as a demonstrating task, we show that the deep memory network significantly outperforms classical text expansion methods with comprehensive experiments on real world data sets.

cs.CL↗

arXiv2017

CirCNN: Accelerating and Compressing Deep Neural Networks Using Block-CirculantWeight Matrices

Large-scale deep neural networks (DNNs) are both compute and memory intensive. As the size of DNNs continues to grow, it is critical to improve the energy efficiency and performance while maintaining accuracy. For DNNs, the model size is an important factor affecting performance, scalability and energy efficiency. Weight pruning achieves good compression ratios but suffers from three drawbacks: 1) the irregular network structure after pruning; 2) the increased training complexity; and 3) the lack of rigorous guarantee of compression ratio and inference accuracy. To overcome these limitations, this paper proposes CirCNN, a principled approach to represent weights and process neural networks using block-circulant matrices. CirCNN utilizes the Fast Fourier Transform (FFT)-based fast multiplication, simultaneously reducing the computational complexity (both in inference and training) from O(n2) to O(nlogn) and the storage complexity from O(n2) to O(n), with negligible accuracy loss. Compared to other approaches, CirCNN is distinct due to its mathematical rigor: it can converge to the same effectiveness as DNNs without compression. The CirCNN architecture, a universal DNN inference engine that can be implemented on various hardware/software platforms with configurable network architecture. To demonstrate the performance and energy efficiency, we test CirCNN in FPGA, ASIC and embedded processors. Our results show that CirCNN architecture achieves very high energy efficiency and performance with a small hardware footprint. Based on the FPGA implementation and ASIC synthesis results, CirCNN achieves 6-102X energy efficiency improvements compared with the best state-of-the-art results.

cs.CV↗

arXiv2017

Probing Direct and Indirect Unitarity Violation in Future Accelerator Neutrino Facilities

The possible existence of light and heavy sterile neutrinos may give rise to direct and indirect unitarity violation of the $3\times3$ lepton mixing matrix respectively. In the current work we study the potential of future accelerator neutrino facilities in probing the unitarity violation effects. Taking DUNE, T2HK and a low-energy Neutrino Factory (LENF) as working examples of future accelerator neutrino facilities, we study the distinct effects of direct and indirect unitarity violation on the discovery reach of the leptonic CP violation and precision measurements of $θ_{23}$ in the three neutrino framework. In addition, constraints on the additional mixing parameters of direct and indirect unitarity violation are also discussed. Finally, we stress that the combination of experiments with different oscillation channels, different neutrino beams and different detector techniques will be an effective solution to the parameter degeneracy problem and give the robust measurement of leptonic CP violation even if the direct and indirect unitarity violation are taken into account.

hep-ph↗

arXiv2017

A Hierarchical Framework of Cloud Resource Allocation and Power Management Using Deep Reinforcement Learning

Automatic decision-making approaches, such as reinforcement learning (RL), have been applied to (partially) solve the resource allocation problem adaptively in the cloud computing system. However, a complete cloud resource allocation framework exhibits high dimensions in state and action spaces, which prohibit the usefulness of traditional RL techniques. In addition, high power consumption has become one of the critical concerns in design and control of cloud computing systems, which degrades system reliability and increases cooling cost. An effective dynamic power management (DPM) policy should minimize power consumption while maintaining performance degradation within an acceptable level. Thus, a joint virtual machine (VM) resource allocation and power management framework is critical to the overall cloud computing system. Moreover, novel solution framework is necessary to address the even higher dimensions in state and action spaces. In this paper, we propose a novel hierarchical framework for solving the overall resource allocation and power management problem in cloud computing systems. The proposed hierarchical framework comprises a global tier for VM resource allocation to the servers and a local tier for distributed power management of local servers. The emerging deep reinforcement learning (DRL) technique, which can deal with complicated control problems with large state space, is adopted to solve the global tier problem. Furthermore, an autoencoder and a novel weight sharing structure are adopted to handle the high-dimensional state space and accelerate the convergence speed. On the other hand, the local tier of distributed server power managements comprises an LSTM based workload predictor and a model-free RL based power manager, operating in a distributed manner.

cs.DC↗

arXiv2017

Joint Mode Selection and Resource Allocation for D2D Communications via Vertex Coloring

Device-to-device (D2D) communication underlaid with cellular networks is a new paradigm, proposed to enhance the performance of cellular networks. By allowing a pair of D2D users to communicate directly and share the same spectral resources with the cellular users, D2D communication can achieve higher spectral efficiency, improve the energy efficiency, and lower the traffic delay. In this paper, we propose a novel joint mode selection and channel resource allocation algorithm via the vertex coloring approach. We decompose the problem into three subproblems and design algorithms for each of them. In the first step, we divide the users into groups using a vertex coloring algorithm. In the second step, we solve the power optimization problem using the interior-point method for each group and conduct mode selection between the cellular mode and D2D mode for D2D users, and we assign channel resources to these groups in the final step. Numerical results show that our algorithm achieves higher sum rate and serves more users with relatively small time consumption compared with other algorithms. Also, the influence of system parameters and the tradeoff between sum rate and the number of served users are studied through simulation results.

cs.IT↗

arXiv2017

Roaming across the Castle Tunnels: an Empirical Study of Inter-App Navigation Behaviors of Android Users

Mobile applications (a.k.a., apps), which facilitate a large variety of tasks on mobile devices, have become indispensable in our everyday lives. Accomplishing a task may require the user to navigate among various apps. Unlike Web pages that are inherently interconnected through hyperlinks, mobile apps are usually isolated building blocks, and the lack of direct links between apps has largely compromised the efficiency of task completion. In this paper, we present the first in-depth empirical study of inter-app navigation behaviors of smartphone users based on a comprehensive dataset collected through a sizable user study over three months. We propose a model to distinguish informational pages and transitional pages, based on which a large number of inter-app navigation are identified. We reveal that developing 'tunnels' between of isolated apps has a huge potential to reduce the cost of navigation. Our analysis provides various practical implications on how to improve app-navigation experiences from both the operating system's perspective and the developer's perspective.

cs.HC↗

arXiv2017

Spectral and Energy Efficiency in Cognitive Radio Systems with Unslotted Primary Users and Sensing Uncertainty

This paper studies energy efficiency (EE) and average throughput maximization for cognitive radio systems in the presence of unslotted primary users. It is assumed that primary user activity follows an ON-OFF alternating renewal process. Secondary users first sense the channel possibly with errors in the form of miss detections and false alarms, and then start the data transmission only if no primary user activity is detected. The secondary user transmission is subject to constraints on collision duration ratio, which is defined as the ratio of average collision duration to transmission duration. In this setting, the optimal power control policy which maximizes the EE of the secondary users or maximizes the average throughput while satisfying a minimum required EE under average/peak transmit power and average interference power constraints are derived. Subsequently, low-complexity algorithms for jointly determining the optimal power level and frame duration are proposed. The impact of probabilities of detection and false alarm, transmit and interference power constraints on the EE, average throughput of the secondary users, optimal transmission power, and the collisions with primary user transmissions are evaluated. In addition, some important properties of the collision duration ratio are investigated. The tradeoff between the EE and average throughput under imperfect sensing decisions and different primary user traffic are further analyzed.

cs.IT↗