Application of BP neural network based on principal component analysis in grain yield prediction

Based on the problem that when BP neural network is used in grain yield prediction, if the input space is too self relevant, the predicting accuracy of BP neural network would drop. This paper introduces the method handled on input variables in advance by the principal component analysis. Comparing with the common BP neural network model, the result indicates that the model of principal component analysis method in BP neural network has the characteristics of higher precision and faster convergence speed.