Research on the improvement of BP neural network model based on genetic algorithm

BP neural network has been widely used in pattern recognition, predictive analysis, control optimization, data mining, etc. Optimizing its structure holds immense importance. For the sake of making BP neural network better adapt to various complex and nonlinear problems, this paper enhances the BP neural network utilizing a genetic algorithm. The BP neural network employs a genetic algorithm to optimize its initial weight and threshold. Hence this model has powerful worldwide search capability of genetic algorithm and the learning ability and adaptation ability of BP neural network, and is capable of efficiently resolving the issue of overfitting or underfitting. The practice results show that the RMSE, MAE, MAPE and R of BP neural network model based on genetic algorithm are 0.19768, 0.16326, 0.015414 and 0.9995, which have been greatly improved in terms of model algorithm and computing speed.