Time Series Prediction of Bank Cash Flow Based on Grey Neural Network Algorithm

For improving the forecasting accuracy of bank cash flow, a combined model based on back propagation (BP) neural network and grey prediction method is put forward based on the merits and demerits of both BP neural network and grey model prediction method. The proposed method has the advantage of two methods and makes up the deficiencies of single model as well. It can efficiently reduce the influence of predicting precision caused by high data fluctuation, and is also capable of enhancing the self-adaptability of forecasting. The accumulation generating operation of grey prediction method is used to transform the original data to generate the accumulated data with better regularity so as to facilitate the neural network modeling and training. By using the function approximation feature of neural network, the prediction of raw bank cash flow data can be realized. The simulation comparison experiments and the results show the BP neural network can revise the GM (1, 1) so that the predictive accuracy of the combined model is higher than individual GM (1, 1).