Modeling and Analyzing of Regional Fluctuation of Cocoon Silk Based on BP Neural Network

Abstract

According to the collected data of the market monthly closing price on dry cocoon and raw silk, predictive modeling and analyzing on the price trend of dry cocoon and raw silk are made based on the related theories and test analysis of BP Artificial Neural Network is carried out. Developed corresponding procedure with Neural Network toolbox under the condition of MATLAB, and then set up BP network relevant prediction models, at last, checked up with examples. At the same time, this study set up corresponding time series forecasting models and made empirical analysis on the basis of Eviews software. The results show that, the two methods both fit for short-term prediction, BP network can achieve human coordination control, the better predictive precision, which supplies an analysis way for silk cocoon market, all of that can be referred to in the future.