Research on Economic Data Analysis and Intelligent Prediction Based on BP Neural Network

Abstract

Prediction is the premise of decision-making, and scientific decision-making can only be made on the basis of correct prediction. Macroeconomic forecasting and decision-making is an important research direction in the field of management science, and an important problem that must be solved in regional and national economic development planning and decision-making. By using the self-learning, self-adapting and nonlinear characteristics of BPNN (BP neural network), economic data analysis and intelligent prediction can be realized. By establishing the evaluation index system of economic system, the data of economic variables are normalized, and then sent to BPNN for training to get the corresponding parameters before prediction. In this paper, PCA (principal component analysis) algorithm and BPNN algorithm are combined, and the PCA algorithm's advantage of dimension reduction and neural network's advantage of nonlinear expression are fully utilized, and the PCA-BPNN prediction model is established, and the algorithm is applied to the analysis of social fixed assets investment data. Compared with the linear prediction method, it is found that PCA-BPNN prediction algorithm has better effect.