PCA-GA-SVM模型的构建及应用研究——沪深300指数预测精度实证分析

Research for Construction and Application of PCA-GA-SVM Model

本文在传统SVM方法的基础上,引入主成分分析方法和遗传算法,构建了PCA-GA-SVM模型,该模型解决了传统SVM方法存在的特征指标相关性、包含惩罚系数和核函数的参数无法动态寻优的问题,最后利用沪深300指数和前五大成份股的日走势数据对该模型进行了验证分析。结果表明,本文所构建的模型,对于沪深300指数和大盘股每日走势的预测精度是很高的,这对于政府管理层面监测股票市场的平稳波动有着很好的应用价值。

This paper extends the traditional SVM method with Principal Component Analysis method and Genetic Algorithm method to construct the PCA - GA - SVM model. The model is in order to solve the traditional SVM method with the problem of the correlation between feature indicators and the dynamic optimization of penalty coefficient C and kernel function＇s parameters. Finaly we use the daily closing price of HuShen 300 Index and component stocks of this index to test the high prediction accuracy of PCA - GA - SVM model. The empirical results show that the model has very effective prediction accuracy for the daily movements of HuShen 300 Index and large capitalization stocks. This result illustrates that the relevant government can take use of the model to monitor the smooth fluctuations in the stock market.