Optimize BP Neural Network Structure on Car Sales Forecasts Based on Genetic Algorithm

Abstract

In order to improve the ability of BP neural network to fit complex functions, we improve the structure of the BP neural network and optimize the weights and thresholds of structure of the BP neural network based on genetic algorithm, then, training the BP neural network model to improve its capability, so, we can apply the model to the automobile sales forecasting system. We compare the prediction accuracy with the traditional BP neural algorithm, it shows that this method obviously fits the data better and has higher prediction accuracy to dates with significant linear correlation