比较BP神经网络和RBF神经网络在基金净值预测中的应用

基金市场的活跃程度直接影响基金净值的变动,市场内部的影响因素具有较强的非线性特征,神经网络模型强大的非线性处理功能能够更为精准地预测基金净值的走势.本文采用BP神经网络和RBF神经网络对华夏成长基金进行实证分析,比较2种方法的预测精度.实证结果表明:RBF神经网络的仿真结果与真实值匹配程度较好,具有更高的预测精度.

Comparison of the application of BP neural network and RBF neural network to the prediction of the net asset value of fund

The active degree of the fund market is directly affecting the net asset value( NAV) of fund. The influencing factors of the internal market have strong nonlinear characteristics,and the strong nonlinear processing functions of the neural network model can predict NAV movement more accurately. This paper uses BP neural network and RBF neural network to make an empirical analysis of Huaxia Growth Fund,comparing two methods of prediction accuracy. The empirical result shows that the results of RBF neural network simulation match well with the real values,and have a higher prediction accuracy.