p.1520
p.1525
p.1529
p.1533
p.1537
p.1541
p.1546
p.1550
p.1554
Application of Immune Algorithm on Input Variable Dimension Reduction in Neural Network Modeling
Abstract:
When the input variables are not mutual independence, it is easy to arise over fitting phenomenon using neural network modeling. So it leads to not only low accuracy of the proposed model but also the long time of modeling .In this paper using immune algorithm to optimize the input variables and selection, The simulation results show that the established model after optimization improves the performance, increases the accuracy and shortens the time of the modeling. Which shows that this algorithm can be effectively reduce the dimension of input variables of neural network.
Info:
Periodical:
Pages:
1537-1540
Citation:
Online since:
September 2013
Authors:
Price:
Сopyright:
© 2013 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: