Research on Two-Sided Combined Forecasting Method of Economic Supply and Demand Based on Big Data Analysis

Aiming at the low accuracy of traditional forecasting methods, this paper studies the combination forecasting of economic supply and demand based on large data analysis. Preprocessing the supply and demand data of historical economy which may be abnormal, wrong and missing. The processed data are input into BP neural network and GM model as input data. Through the training of BP neural network and the processing of GM model, set parameters. The annealing algorithm is used to calculate the combination weight. According to the combination weight, the BP neural network and GM model are linearly combined to complete the design of the two-sided combination forecasting method for economic supply and demand. The experimental results show that the prediction accuracy of the traditional prediction method is about 6/7 of that of the proposed method, and the prediction accuracy is significantly improved.