Prediction of the logistics demand for Heilongjiang province based on radial basis function algorithm

In order to predict the logistics demand accurately, neural network model based on radial basis function (RBF) algorithm is used. The data of 1991 to 2005 are chosen as training samples. The samples from 2005 to 2008 as input variant are used to test the data from 2006 to 2009. The results shows that the maximal relative error is 2.14% (<;4%). RBF network model through training can predict the logistics demand exactly with better generalization in addition. The results showed that the established neural network model have both satisfying fitting and predicting precision. Conclusions can be drawn that the model is more accurate. It has certain practical value according to the establishment of RBF neural network model for predicting logistics demand.