检索结果-内蒙古大学图书馆

A comprehensive reconfigurable computing approach to memory wall problem of large graph computation

JOURNAL OF SYSTEMS ARCHITECTURE 2016年 70卷 59-69页

作者： Wang, Xu Zhu, Yongxin Huang, Linan Shanghai Jiao Tong Univ Sch Microelect 800 Dongchuan Rd Shanghai 200240 Peoples R China Natl Univ Singapore Sch Comp Singapore Singapore

Graph computation problems that exhibit irregular memory access patterns are known to show poor performance on multiprocessor architectures. Although recent studies use FPGA technology to tackle the memory wall problem of graph computation by adopting a massively multi-threaded architecture, the performance is still far less than optimal memory performance due to the long memory access latency. In this paper, we propose a comprehensive reconfigurable computing approach to address the memory wall problem. First, we present an extended edge-streaming model with massive partitions to provide better load balance while taking advantage of the streaming bandwidth of external memory in processing large graphs. Second, we propose a two-level shuffle network architecture to significantly reduce the on chip memory requirement while provide high processing throughput that matches the bandwidth of the external memory. third, we introduce a compact storage design based on graph compression schemes and propose the corresponding encoding and decoding hardware to reduce the data volume transferred between the processing engines and external memory. We validate the effectiveness of the proposed architecture by implementing three frequently-used graph algorithms on ML605 board, showing an up to 3.85 x improvement in terms of performance to bandwidth ratio over previously published FPGA-based implementations. (C) 2016 Elsevier B.V. All rights reserved.

关键词： Embedded architecture FPGA Graph computing Hardware implementation Memory wall problem

来源：评论

学校读者我要写书评

暂无评论

Verifiable computation for Randomized Algorithm 17th

Verifiable Computation for Randomized Algorithm

引用

17th international conference on Cryptology in India (INDOCRYPT)

作者： Liu, Muhua Wu, Ying Xue, Rui Chinese Acad Sci Inst Informat Engn State Key Lab Informat Secur Beijing 100093 Peoples R China Univ Chinese Acad Sci Beijing 100049 Peoples R China

ISBN: (纸本)9783319498904;9783319498898

Verifiable computation enables a computationally weak client to delegate the difficult computation to a more powerful cloud server. When the client receives the returned result, it can verify the correctness of the result. As a new computing model, it has been widely studied. But, the previous works on verifiable computation was restricted to the case where the target function was deterministic. In this paper, we present a novel definition for verifiable computation supporting randomized algorithms, which allows a resource constrained client to outsource the computation of a randomized algorithm to an untrusted server. We consider a new setting where the random string, as the random input of the randomized algorithm, is generated by the cloud server. In our security definition, it must guarantee that the server can not tamper with the randomness which means that the generated string is indistinguishable with a truly random string. Our construction is based on indistinguishability obfuscation, constrained verifiable random function and functional pseudorandom function.

关键词： Verifiable computation Randomized algorithms Indistinguishability obfuscation

来源：评论

学校读者我要写书评

暂无评论

Parallel image dehazing algorithm based on GPU using fuzzy system and hybird evolution algorithm

Parallel image dehazing algorithm based on GPU using fuzzy s...

引用

ACIS international conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing (SNPD)

作者： Che-Lun Hung Ren-You Yan Hsiao-Hsi Wang Department of Computer Science and Communication Engineering Providence University Taichung Taiwan Department of Computer Science and Information Management Providence University Taichung Taiwan

Haze is an atmospheric phenomenon that fogs the visibility of the scenes. Removing the haze has been an important issue in image processing technologies. Many image dehazing technologies with evolution algorithms have been proposed to remove the fog in the image. However, these algorithms usually are compute-intensive. In this paper, we propose a parallel hybrid evolution algorithm based on GPU to enhance the computational performance. In traditional evolution algorithms, the calculation of fitness function occupies the most of the computation time. In the proposed method, we implement this part on GPU by using CUDA framework to reduce the computational load. the experiment results show that the proposed method can remove the haze efficiently and successfully.

关键词： Graphics processing units Fuzzy systems Evolutionary computation Signal processing algorithms Algorithm design and analysis Genetic algorithms Entropy

来源：评论

学校读者我要写书评

暂无评论

Dimension Reduction EA for Portfolio Replication Problem: Search Space Change by Fixing Important Variables

Dimension Reduction EA for Portfolio Replication Problem: Se...

引用

international Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Yukiko Orito Yoshiko Hanada Department of Economics Hiroshima University Hiroshima Japan Faculty of Engineering Science Kansai University Osaka Japan

ISBN: (纸本)9781509026791

the portfolio optimization problem is viewed as one of high-dimensional problems. It is difficult to find an optimal solution in the high-dimensional problem because evolutionary algorithms have to determine many design variables for a solution. In this paper, we propose a dimension reduction approach in an evolutionary algorithm for this problem. Our approach changes the search space by the dimension reduction which fixes important design variables. In the numerical experiments, we show that our approach is very useful for the high-dimensional portfolio replication problem.

关键词： Portfolios Sociology Statistics Decision making Benchmark testing Companies Evolutionary computation

来源：评论

学校读者我要写书评

暂无评论

Evolutionary Algorithmic Parameter Optimization of MOEAs for Multiple Multi-objective Problems

Evolutionary Algorithmic Parameter Optimization of MOEAs for...

引用

international Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Motoaki Kakuguchi Minami Miyakawa Keiki Takadama Hiroyuki Sato Graduate School of Information and Engineering Sciences University of Electro-Communications Tokyo Japan

ISBN: (纸本)9781509026791

Nowadays, algorithmic studies of multi-objective evolutionary algorithms (MOEAs) are flooded with too many search algorithms. Each MOEA has its own expert problem domain. To clarify not only the optimal MOEA and its parameters for each of multiple multi-objective optimization problems (MOPs) but the robust MOEAs for multiple MOPs, this work proposes a meta-MOEA framework to search the Pareto optimal algorithmic parameters for multiple MOPs. In this work, we use two DTLZ2 benchmark problems with 2 and 4 objectives and optimize the base algorithm, the crossover rate and its parameter, the mutation rate and its parameter for the both DTLZ2 problems by the meta-MOEA. the experiment results show that the optimal algorithmic parameters for each of two DTLZ2 problems are different and the robust algorithmic parameters for both problems can be obtained by the meta-MOEA framework.

关键词： Optimization Sociology Statistics Indexes Evolutionary computation Benchmark testing Search problems

来源：评论

学校读者我要写书评

暂无评论

Effects of Different Implementations of a Real Random Number Generator on the Search Behavior of Multiobjective Evolutionary algorithms

Effects of Different Implementations of a Real Random Number...

引用

international Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Takahiro Funakoshi Yusuke Nojima Hisao Ishibuchi Graduate School of Engineering Osaka Prefecture University Osaka Japan

ISBN: (纸本)9781509026791

We often notice that different experimental results of the same algorithm were reported in the literature. Usually this is because computational experiments were performed under different settings. However, even when almost the same settings are used, similar results are not always obtained. Especially when we use our own implementation, it is very difficult to obtain similar experimental results to the reported ones in the literature. Of course, due to the stochastic nature of an evolutionary algorithm, its experimental results can be different in each run. However, its average results over many runs should be similar. Unfortunately, this is not always the case. We often encounter the situation where our own experimental results are clearly different from the reported ones in the literature. In this paper, we report our investigation results on the following question: why totally different results of MOEA/D were often reported in the literature. More specifically, we show that totally different results can be obtained when we generate random real numbers in a slightly different manner such as the normalization of integers in [0, 253 - 1] by dividing them by 253 or (253 - 1).

关键词： Generators Optimization Evolutionary computation Approximation algorithms Indexes Intelligent systems Genetics

来源：评论

学校读者我要写书评

暂无评论

Probabilistic Model-Based Multistep Crossover for Genetic Programming

Probabilistic Model-Based Multistep Crossover for Genetic Pr...

引用

international Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Kohei Matsumura Yoshiko Hanada Keiko Ono Graduate School of Science and Engineering Kansai University Osaka Japan Faculty of Engineering Science Kansai University Osaka Japan Department of Electronics and Informatics Ryukoku University Shiga Japan

ISBN: (纸本)9781509026791

Deterministic Multistep crossover fusion (dMSXF) is one of promising crossover methods of a tree-based genetic programming. dMSXF performs a multistep local search from a parent in the direction approaching the other parent. In the local search, neighborhood solutions are generated by operators based on a replacement, an insertion and a deletion of nodes to combine both parents' small trait step by step. Due to this mechanism, dMSXF can generate a wide variety of solution between parents. However, some random nodes are inserted or deleted in the solution at each step of the local search to satisfy constraints, which sometimes cause the generation of undesirable neighborhood solutions. In this paper, we introduce a probabilistic model constructed by the search information to the generation of neighborhood solutions in order to improve the search efficiency of dMSXF. the search performance of the proposed method is evaluated on symbolic regression problems and the Santa Fe Trail problem.

关键词： Probabilistic logic Search problems Genetic programming Evolutionary computation Intelligent systems Electronic mail Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Creation of Warning Sound by Vote of Multiple Users Based on Interactive Differential Evolution: Discussion toward Effective IECs Creating of Media Contents Suited to Multiple Users

Creation of Warning Sound by Vote of Multiple Users Based on...

引用

international Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Makoto Fukumoto Department of Computer Science and Engineering Fukuoka Institute of Technology Fukuoka Japan

ISBN: (纸本)9781509026791

Interactive Evolutionary computation (IEC) is a method for creating media contents suited to user's feelings. IEC is originally used for personal objectives. Recent IEC studies apply it to common objectives between multiple users. For obtaining fitness value of multiple users, the previous IEC studies include interesting schemes, vote, parallel distributed method, and summarization of multiple fitness values. Target of these previous IEC studies was visual contents. this study proposes a method for creating sound contents suited to multiple users' feelings based on Interactive Differential Evolution (IDE). In the proposed method, multiple users' subjective evaluation is obtained by vote for paired solution candidates: the solution candidate gathered larger number of votes retains as a solution candidate in the next generation. the efficiency of the proposed IDE was investigated through listening experiment. Target of creation was "warning sound". the target and parameters of IDE were set by referring to the author's previous study. In the experiment 1, the subjects searched better warning sounds with vote. As result, slow convergence of searching space was observed. In the experiment 2, the subjects evaluated two sign sounds created in the experiment 1. No significant increase in fitness value was observed. these poor results might come from summarization of multiple users' feeling as the selection. Based on the results, improvement method is discussed.

关键词： Media IEC Music Sociology Statistics Evolutionary computation Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

17th international Workshop on Advanced Computing and Analysis Techniques in Physics Research (ACAT 2016)

引用

Journal of Physics: conference Series 2016年第1期762卷

Preface the 2016 version of the international Workshop on Advanced Computing and Analysis Techniques in Physics Research took place on January 18-22, 2016, at the Universidad Técnica Federico Santa Maria -UTFSM- in Valparaiso, Chile. the present volume of IOP conference Series is devoted to the selected scientific contributions presented at the workshop. In order to guarantee the scientific quality of the Proceedings all papers were thoroughly peer-reviewed by an ad-hoc Editorial Committee with the help of many careful reviewers. the ACAT Workshop series has a long tradition starting in 1990 (Lyon, France), and takes place in intervals of a year and a half. Formerly these workshops were known under the name AIHENP (Artificial Intelligence for High Energy and Nuclear Physics). Each edition brings together experimental and theoretical physicists and computer scientists/experts, from particle and nuclear physics, astronomy and astrophysics in order to exchange knowledge and experience in computing and data analysis in physics. three tracks cover the main topics: Computing technology: languages and system architectures. Data analysis: algorithms and tools. theoretical Physics: techniques and methods. Although most contributions and discussions are related to particle physics and computing, other fields like condensed matter physics, earth physics, biophysics are often addressed in the hope to share our approaches and visions. It created a forum for exchanging ideas among fields, exploring and promoting cutting-edge computing technologies and debating hot topics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Novel Fast Approach for Convolutional Networks with Small Filters Based on GPU 17

A Novel Fast Approach for Convolutional Networks with Small ...

引用

2015 IEEE 17th international conference on High Performance Computing and Communications (HPCC)

作者： Jiang, Wenbin Chen, Yiming Jin, Hai Luo, Bin Chi, Ye Huazhong Univ Sci & Technol Sch Comp Sci & Technol Serv Comp Technol & Syst Lab Cluster & Grid Comp Lab Wuhan 430074 Peoples R China

ISBN: (纸本)9781479989379

Recently, convolutional networks have achieved great successes in the field of computer vision. In order to improve the efficiency of convolutional networks, large amount of solutions focusing on training algorithms and parallelism strategies have been proposed. In this paper, a novel algorithm based on look-up table is proposed to speed up convolutional networks with small filters by applying GPU. By transforming multiplication operations in the convolution computation to some table-based summation operations, the overhead of convolution computation can be reduced largely. the process of creating table and looking up table is very appropriate for parallelization on GPU. Experiment results show that the proposed approaches can improve the speed of convolution computation by 20%-30%, compared with state-of-the-art existing works.

关键词： Deep Learning Convolutional Network Parallel computation GPU

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：