信息与控制2016，Vol.45Issue(6)：735-741,758,8.DOI:10.13976/j.cnki.xk.2016.0735
增量型极限学习机改进算法
Improvement Algorithm of an Incremental Extreme Learning Machine
摘要
Abstract
Given that the input weights and threshold of hidden layer neurons are obtained randomly,in the training process of an incremental extreme learning machine (Ⅰ-ELM),the output weights of some hidden layer neurons may be too small to contribute effectively to the network output.This causes the neurons to be invalid.This problem not only makes the network more complicated,but also reduces the stability of the network.To deal with this issue,we propose in this study an improved method that adds an offset to hidden layer output of Ⅰ-ELM (Ⅱ-ELM).Then,we analyze and prove the existence of the offset.Finally,the validity of Ⅱ-ELM is verified by simulationand comparison with the Ⅰ-ELM in classification and regression problems.关键词
增量型极限学习机/无效神经元/算法改进/网络的稳定性Key words
incremental extreme learning machine (Ⅰ-ELM)/invalid neurons/algorithm improvement/network stability分类
信息技术与安全科学引用本文复制引用
宋绍剑,向伟康,林小峰..增量型极限学习机改进算法[J].信息与控制,2016,45(6):735-741,758,8.基金项目
国家自然科学资助基金(61364007) （61364007）
广西科学研究与技术开发计划项目(桂科攻14122007-33) （桂科攻14122007-33）
南宁市科学研究与技术开发计划项目(20141050) （20141050）
