Research On Time Series Prediction Method Of Bank Cash Flow And Simulation Comparison

Due to the rapid growth of China’s commercial banks cash business,in order to improve the accuracy and real-time of all kinds of information in the cash business, and solve the problem which accuracy and stability is not high of the data linkage between cash inventory forecasting and cash management information in the commercial bank.In this paper, the following methods are proposed toresearch ontime series prediction methods of China’s commercialbankscash flow.The specific works are as follows:Firstly, the statistical time series forecasting methods are used which the forecasting methodsare adopted commonly at present.Including the moving average method, the exponential smoothing prediction method, and the autoregressive moving average hybrid model(ARMA model), finishing cash flow time series forecasting based on the market data of a commercial bank. And through a large number of simulation experiments and comparative experiments, verifying the effectiveness of the proposed methods.Secondly, a combined model based on BP neural network and greyprediction methods are proposed to realize the bank cash flow time series prediction, which can gather advantage of the two methods and make up the deficiencies of singlemodel aswell. This combined method can efficiently reduce the influence of predicting precise caused by high data fluctuation and also capable of enhancing the self-adaptability of forecasting. We utilize the accumulation generating operation ofgrey prediction to transform the original data to generate these 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 data and Simulation verification can be realized.Finally, a hybrid learning algorithm is proposed based on adaptive population activity particle swarm optimization(APAPSO) algorithm combined with the least squares method(LMS) to optimize the adaptive network-based fuzzy inference system(ANFIS) model parameters. Through the introduction of metric function of population diversity to ensure the diversity of population and the introduction of adaptive changes in inertia weight, the optimization ability of the PSO algorithm is improved, which avoids the premature convergence problem of PSO algorithm.The large simulation comparison experiments are carried out to verify the effectiveness of the proposed improved algorithm.In conclusion, simulation results show that the above methods can achieve better forecasting effect for the bank’s cash flow time seriesprediction, which provides a good data base and decision basis for making business plan, and has a good reference value.