Prediction of the return of common fund through General Regression Neural Network

Abstract

As the General Regression Neural Network can not only deal with issues of classification can be used for continuous variables in the forecast. In recent years, there have been many relevant documents used successfully in General Regression Neural Network for the financial sector forecasts. In this paper, has also adopted the General Regression Neural Network for the prediction on the net value of the domestic mutual fund and evaluation on the return of the investment. To start the analysis, pick out 53 balance fund for analysis on its investment performance by using gray relational analysis, the selection of good investment performance of mutual funds as an investment target. By using General Regression Neural Network model to set up the prediction model and using grey prediction and multiple regression model to do the comparison analysis on the accuracy of the prediction and the return rate. By using General Regression Neural Network to predict and the return rate is better than using grey prediction and multiple regression model, base on the evaluation of the 5 index number of the performance management, 5 Group interactive data validation map and the return of investment.

Keywords and phrases:

• Grey relational analysis
• grey prediction
• general regression neural network
• multiple regression
• genetic algorithm