Postgraduate entrant and employment forecasting using modified BP neural network with PSO

Abstract

It is hard to train the influence variables and to forecast the complex problems due to the time series. Recently the neural network method has been successfully employed to solve the forecasting problem. In this paper, an approach that integrate modified BP neural network optimized with particle swarm optimization algorithm (MBPPSO) is proposed which applied to forecast postgraduate entrant and employment problem. It introduces particle swarm optimization algorithm to optimize the initial weights of the BP neural network, which effectively improve velocity of convergence BP neural network. Moreover, the adaptive adjust learn strategy is introduced to avoid acutely shake of train and decrease the bias error. The experiment results show MBPPSO can achieve reasonable forecast result. key words-particle swam optimization, BP neural network， postgraduate entrant and employment forecasting I.