基于PSO算法优化RBF神经网络的基金净值预测研究

为了更精确地对基金净值进行预测,针对基金净值变化具有非线性和随机性等特点,提出基于粒子群优化RBF神经网络的基金净值预测模型。利用具有全局寻优的PSO算法对RBF神经网络的参数进行优化,并用经PSO算法优化的RBF神经网络对基金净值进行预测分析。仿真实验结果表明:与使用BP神经网络和RBF神经网络的基金价格预测方法相比较,PSO算法优化的RBF神经网络能够准确地预测基金价格的变化趋势,具有较高的预测精度,对于用户选择基金有着非常重要的意义。

Research on Fund Net Value Prediction Based on RBF Neural Network Optimized by PSO Algorithm

In order to predict fund net value more accurately, the forecasting model of fund net value based on RBF neural network optimized by particle swarm optimization algorithm is proposed for the nonlinear and stochastic characteristics of fund net value. The parameters of RBF neural network are optimized by using PSO algorithm with global optimization and the fund net value is predicted and analyzed by using RBF neural net- work optimized by PSO algorithm. Simulation experimental results show that, compared with the fund predic- tion methods using BP neural network and RBF neural network,RBF neural network optimized by PSO algo- rithm can more accurately predict the changing trend of fund prices, which has higher prediction accuracy and is very significant to select the fund for the users.