Research on inventory forecasting based on BP neural network

Abstract

Accurate prediction of inventory can improve inventory value, speed up the capital turnover and improve the profitability of enterprises. It has become a hot research topic in recent years. Taking the business process of agricultural enterprise as the research background. Using multiple imputation methods to optimize the BP (Back Propagation) neural network inventory forecasting model, and combined with the demand of the inventory business personnel and functional units, we propose a novel inventory prediction method based on BP neural network. The method takes the four factors of conventional, policy, drug resistance and market as the BP neural network to predict the inventory requirements of the input layer samples. When the input layer lack of sample data, this method uses multiple imputation method for missing data for each value instead of fetching the collection, by exploring each value instead of the prediction error trend under the combination to find a better value. After data filtering and missing data interpolation, the sample data is input into the BP neural network prediction model. The simulation results show that the proposed method has less error and has the advantages and advantages than the existing methods.