Research on RBM training algorithm with dynamic gibbs sampling

Fei Li, Xiao Guang Gao, Kai Fang Wan

科研成果: 期刊稿件文章同行评审

12 引用 (Scopus)

摘要

Currently, most algorithms for training restricted Boltzmann machines (RBMs) are based on the multi-step Gibbs sampling. This article focuses on the problems of sampling divergence and the low training speed associated with the multi-step Gibbs sampling process. Firstly, these problems are illustrated and described by experiments. Then, the convergence property of the Gibbs sampling procedure is theoretically analyzed from the prospective of the Markov sampling. It is proved that the poor convergence property of the multi-step Gibbs sampling is the main cause of the sampling divergence and the low training speed when training an RBM. Furthermore, a new dynamic Gibbs sampling algorithm is proposed and its simulation results are given. It has been demonstrated that the dynamic Gibbs sampling algorithm can effiectively tackle the issue of sampling divergence and can achieve a higher training accuracy at a reasonable expense of computation time.

源语言英语
页(从-至)931-942
页数12
期刊Zidonghua Xuebao/Acta Automatica Sinica
42
6
DOI
出版状态已出版 - 1 6月 2016

指纹

探究 'Research on RBM training algorithm with dynamic gibbs sampling' 的科研主题。它们共同构成独一无二的指纹。

引用此