Improved Implementation of Expectation Maximization Algorithm on Graphic Processing Unit

被引：0

作者：

Jing, Si-Yuan ^{[1
]}

Sun, Rui ^{[1
]}

Xie, Chun-Ming ^{[1
]}

Jin, Peng ^{[1
]}

Liu, Yi ^{[1
]}

Liu, Cai-Ming ^{[1
]}

机构：

[1] Leshan Normal Univ, Sch Comp Sci, Leshan 614000, Peoples R China

来源：

CHINESE LEXICAL SEMANTICS, CLSW 2018 | 2018年 / 11173卷

关键词：

Word alignment; GPU; Parallel computation; Expectation-maximization algorithm; CUDA;

D O I：

10.1007/978-3-030-04015-4_54

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In our previous work, an efficient implementation of Expectation-Maximization (EM) algorithm using CUDA has been proposed for high-speed word alignment. The proposed algorithm can gain a 16.8-fold speedup compared to a multi-thread algorithm and a 234.7-fold speedup compared to a sequential algorithm on a modern graphic processing unit (GPU). In this paper, we try to improve the algorithm to achieve better performance. Through analysis of the previous algorithm, we find that two places in "E" step (expectation calculation) are unreasonably designed. An improved CUDA implementation of the EM algorithm is proposed in this paper. Experimental results show that the new algorithm can improve the speed of expectation calculation by 29.4%.

引用

页码：623 / 629

页数：7

共 9 条

[1] [Anonymous], 2010, PROGRAMMING MASSIVEL
[2] [Anonymous], 2010, Statistical Machine Translation
[3] Brown P. F., 1993, Computational Linguistics, V19, P263
[4] Gao Q, 2008, ACL-08 HLT-Softw. Eng. Testing, Qual. Assur. Nat. Lang. Process, P49, DOI 10.3115/1622110.1622119
[5] Jing SY, 2016, 2016 17TH INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES (PDCAT), P189, DOI [10.1109/PDCAT.2016.49, 10.1109/PDCAT.2016.050]
[6] Och FJ, 2000, 38TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, PROCEEDINGS OF THE CONFERENCE, P440
[7] Optimization Principles and Application Performance Evaluation of a Multithreaded GPU Using CUDA
Ryoo, Shane
Rodrigues, Christopher I.
Baghsorkhi, Sara S.
Stone, Sam S.
Kirk, David B.
Hwu, Wen-mei W.
[J]. PPOPP'08: PROCEEDINGS OF THE 2008 ACM SIGPLAN SYMPOSIUM ON PRINCIPLES AND PRACTICE OF PARALLEL PROGRAMMING, 2008, : 73 - 82
[8] Satish N, 2009, INT PARALL DISTRIB P, P257
[9] Xiao Tong, 2012, Proceedings of the 52th Annual Meeting of the Association for Computational Linguistics, System Demonstrations, P19

← 1 →