Study on primary product logistics: demand prediction based on neural network theory

Primary product logistics shares the challenges of other logistical problems, but also possesses many unique features which preclude the application of usual methods of the logistics of primary products. In particular, it is not possible to accurately forecast demand. To overcome the limitations of single logistics demand forecasting techniques and the difficulties in primary products logistics that exist currently, this paper reports the use of neural network theory to establish a predictive model of the demand in primary products logistics based on a back-propagation (BP) neural network. The BP Algorithm used in the learning process includes two processes: forward computing of data stream and backward propagation of error signals, which make the output vector closer to the expected output vectors by continuous adjusting of weights, thus improving the accuracy of the logistics forecasting. Primary products demand and example Analysis verify the accuracy of this BP neural network based prediction model for primary product demand.