摘要
回声状态网络是一种新型的递归神经网络,近年来引起诸多学者的关注,对回声状态网络的研究也逐步深入。系统介绍回声状态网络的网络结构、学习过程和主要参数,详细论述几种比较热门的改进算法以及回声状态网络目前的应用研究情况,归纳并且比较该算法与BP神经网络、SVM支持向量机等预测模型各自的优缺点以及适用范围,最后总结回声网络目前存在的问题以及未来的研究方向。
The echo state network（ ESN） is a new kind of recurrent neural network which draws the attention of many researchers in recent years,and the research on the ESN is also gradually in-depth. In this paper,the ESN network structure,learning process and major parameters of reservoir are introduced. Several improved algorithms and the applications of the ESN at present are discussed in detail. Then the advantages,disadvantages and applied fields of ESN,BP neural network and support vector machine（ SVM） prediction model are compared. Finally the existing problems in current research are summarized and future research directions are proposed.
出处
《成都信息工程学院学报》
2015年第6期546-550,共5页
Journal of Chengdu University of Information Technology
基金
四川省科技支撑计划资助项目(2014SZ0207)
关键词
回声状态网络
储备池
递归神经网络
echo state network
reservoir
recurrent neural network
