Hits:

Date of Publication:2022-10-10

Journal:系统工程

Volume:30

Issue:10

Page Number:112-117

ISSN No.:1001-4098

Abstract:For the trend of falling into local minimum of BP algorithm and the
   premature convergence of simple genetic algorithm, an improved wavelet
   neural network model based on immune genetic and BP algorithm (IGA-BP)
   is presented, which takes full advantage of the global optimization
   ability of genetic algorithm and the local search ability of BP
   algorithm. BP algorithm is incorporated into immune genetic algorithm to
   optimize the parameters of wavelet neural network by local synchronized
   searching in the solution space that the genetic algorithm has searched
   out. The improved algorithm is used to corporate performance evaluation
   and the results show that the algorithm has higher accuracy than the
   genetic algorithm and the BP algorithm.

Note:新增回溯数据