极限学习机(ELM)网络结构调整方法综述  

A Summary of Network Structure Adjustments of Extreme Learning Machine

作　　者：翟敏[1] 张瑞[1] 王宇[1] 

机构地区：[1]西北大学数学系,西安710127

出　　处：《西安文理学院学报（自然科学版）》2014年第1期1-6,共6页Journal of Xi’an University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61075050)

摘　　要：从原始ELM算法和增长结构ELM算法(I-ELM)的基本思想与基本理论出发,分析其优点与不足,概括基于不同角度所改进的网络结构调整方法,包括结构增长型算法、结构递减型算法和自适应型算法三大方面.最终对极限学习机(ELM)中网络结构的各种调整方法进行综述,为相关研究者提供该研究方向的发展历史和最新结果.We give an analysis of the advantages and disadvantages of original ELM and I-ELM algorithms by addressing their basic theories. We have examined the improved methods of network structure adjustment from various perspectives,which include structural growth algorithm,structural decline algorithm and self- adapting algorithm. We also give a review of the various adjustment methods of network structure. The past research and recent findings reviewed are supposed to be of benefit for the researchers.

关 键 词：ELM算法 网络结构 隐节点 

分 类 号：O29[理学—应用数学]