THE APPLICATION OF ERROR STATISTICS IN NEURAL NETWORK

摘要：

This paper analyzes error distribution between the predictive value and the actual value with the application of Neural Network. Based on Probability theory and Rough set theory, how to revise the predictive value, give more reliable predictive interval and give the degree of confidence are discussed.

关键词： Neural Network Error Statistics Rough set theory degree of confidence

作者: He Haiqian Peng Dewei Li Donghui

作者单位: Department of Computer Science and Technology, Wuhan University of Technology, yujiatou campus Wuhan, Hubei 430063.P.R.China

会议类型: 国际会议

会议名称: 第九届分布式计算及其应用国际学术研讨会

会议地点: 香港

会议语种:英文

页码: 647-649

在线出版日期: 2010-08-12（万方平台首次上网日期，不代表论文的发表时间）