基于非参数回归的遗传神经网络集成股市预测研究

利用遗传算法改进神经网络集成个体的连接结构和初始连接权值,利用主成分分析法提高集成个体差异度,形成一组优良的神经网络集成个体,利用非参数回归生成集成结论,求出非线性时序函数的全局最优解,随即建立新型的基于非参数回归的遗传神经网络集成股市预测模型.仿真结果表明,该模型预测精度高,可操作性强,具有一定实用性.

Stock Market Forecasting Model Based on Nonparametric Regression Evolving Genetic Algorithm Neural Network Ensembles

The connection architecture of network and the connection weights of network are optimized by the implementation of genetic algorithms operation.The best individual neural network could be generated by the principal component analysis（PCA） technology.The ensemble strategy is carried out by nonparametric regression.The global optimal solution of nonlinear time series function was obtained.The novel non-parametric regression（NR） based on genetic neural network ensemble was established for the forecasting model of the stock market.Stimulation testing revealed that the NR ensemble model proposed could be used as an alternative forecasting tool for a meteorological application in achieving greater forecasting accuracy and improving prediction quality.