Nonlinear time series forecasting of time-delay neural network embedded with Bayesian regularization

Based on the idea of nonlinear prediction of phase space reconstruction, this paper presents a time-delay BP neural network model, whose generalization capability is improved by Bayesian regularization. Furthermore, the model is applied to forecast the imp and exp trades in one industry. The results show that the improved model has excellent generalization capabilities, which not only learns of the historical curve, but efficiently predicts the trend of business. Comparing with other forecasts, we draw a conclusion that nonlinear forecast can not only focus on data combination and precision improvement, but also vividly reflect the nonlinear characteristic of the forecasting system. While analyzing the forecasting precision of the model, we give a model judgment by calculating the nonlinear characteristic value of the combined serial and original serial, which proves that the forecasting model can reasonably 'catch' the dynamic characteristic of the nonlinear system which produces the origin serial.