Improved Extreme Learning Machine based on the Sensitivity Analysis

International Conference on Robotics and Mechantronics

机械制造;无线电电子学;计算机科学

Cui, Licheng^1,2,3 ; Zhai, Huawei^3 ; Wang, Benchao^1 ; Qu, Zengtang^1

Liaoning Police College, Dalian

116036, China^1

Faculty of Electronic Information and Electrical Engineering, Dalian University of Technology, Dalian

116024, China^2

Information Science and Technology, Dalian Maritime University, Dalian

116026, China^3

关键词: Extreme learning machine;    Hidden nodes;    Importance;    Learning error;    Learning time;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/320/1/012015/pdf
DOI  :  10.1088/1757-899X/320/1/012015

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Extreme learning machine and its improved ones is weak in some points, such as computing complex, learning error and so on. After deeply analyzing, referencing the importance of hidden nodes in SVM, an novel analyzing method of the sensitivity is proposed which meets people's cognitive habits. Based on these, an improved ELM is proposed, it could remove hidden nodes before meeting the learning error, and it can efficiently manage the number of hidden nodes, so as to improve the its performance. After comparing tests, it is better in learning time, accuracy and so on.