A method of improving generalization ability for neural network based on genetic algorithm

摘要

In order to solve the problem that neural network learns well but predicts badly, the genetic algorithm was adopted to optimize the neural network. The LM-BP neural network learns very well, and it is sensitive to the initial weights and thresholds. Then its initial weights and thresholds were selected by genetic algorithm. So the method of improving generalization ability for neural network based on genetic algorithm was proposed. By example analysis, compared with the method that the initial weights and thresholds were selected randomly, the neural network optimized by genetic algorithm has very high fitting precision and testing accuracy. The new method can greatly improve the generalization ability of neural network.