Forecasting methods and application of regional logistics demand based on wavelet neural network

Abstract

Wavelet neural network is a neural network combining the wavelet theory with neural network theory, which avoids nonlinear optimization problems such as blindness of the design of BP neural network structure and local optimum, greatly simplifying the training. The use of wavelet neural networks to forecast regional logistics demand provided an important reference for regional logistics systematic planning and the rational allocation of logistic resources. Therefore, the use of regional economic indicators to forecast regional logistics demand had strong feasibility and promotes the coordinated development between regional logistics industry and regional economy. The model reveals nonlinear mapping relationship between the regional economy and regional logistics demand and provides a new idea and method for the regional logistics demand forecasting.