A modified differential evolution algorithm and its application in the training of BP neural network

Author

Gao, Yuelin ; Liu, Junmin

Author_Institution

Sch. of Inf. & Comput. Sci., Univ. of North Nat. Univ., Yinchuan

fYear

2008

fDate

2-5 July 2008

Firstpage

1373

Lastpage

1377

Abstract

The paper is given a new modified differential evolution (MDE) algorithm in which a novel mutation operator is introduced. The MDE algorithm can obtain a good balance between global search and local search and was applied in BP neural network training. The numerical results demonstrate that the new MDE algorithm has the abilities of good global search and faster convergence speed and higher convergence accuracy. It can overcome the disadvantages of the traditional BP algorithm and reduce the training time and improve the training accuracy.

Keywords

backpropagation; evolutionary computation; mathematical operators; neural nets; search problems; BP neural network training; convergence; global search; local search; modified differential evolution algorithm; mutation operator; Approximation algorithms; Artificial neural networks; Evolution (biology); Feedforward neural networks; Feeds; Genetic mutations; Model driven engineering; Neural networks; Neurons; Recurrent neural networks; BP neural network; application; differential evolution algorithm; mutation operator;

fLanguage

English

Publisher

ieee

Conference_Titel

Advanced Intelligent Mechatronics, 2008. AIM 2008. IEEE/ASME International Conference on

Conference_Location

Xian

Print_ISBN

978-1-4244-2494-8

Electronic_ISBN

978-1-4244-2495-5

Type

conf

DOI

10.1109/AIM.2008.4601862

Filename

4601862