A distributed data high-frequency storage method based on neural network

Through deep research on high frequency data storage problems and neural network technologies, a method of solving high frequency data storage problems is proposed, The method applies perceptron neural network and BP neural network technologies on distributed data high-frequency storage. It uses perceptron neural network to set up classification model which decides success or failure of high-frequency data storage, and uses BP neural network to predict the size of each client input data after increasing new clients in the distributed system. To verify the effectiveness of the method, it uses the actual input data of multiple clients as test and training data, and compares with exponential smoothing method. Simulation results show that the method solves the instability problems of distributed data high-frequency storage, and has good comprehensive performance.