Forecast of Yellow River Sandiness Based on RBF Neural Network

摘要

Through the neural network to realize the forecast of Yellow River sandiness by means of carrying on the analysis and the discussion. Through selecting the RBF neural network, and using the improvement K- means cluster algorithm, to dynamic determinate the RBF neural network center, meanwhile using the least squares method to adjust the weight of RBF the neural network. By analyzing and processing the Yellow River sandiness data about Lanzhou and Qingtongxia and Sanmenxia, and the result of an example shows that the prediction and forecasting precision are preferable. It may be realize to Yellow River sandiness future forecast. This forecast model has the good forecast effect to the exchange rate.