基于CFOA优化模糊支持向量机的上证综指趋势预测研究  

Research on Trend Forecast of Shanghai Composite Index Based on CFOA Optimized Fuzzy Support Vector Machine

作　　者：闫雅雯 YAN Yawen(College of Economics and Management,Xianyang Normal University,Xianyang 712000,China)

机构地区：[1]咸阳师范学院经济与管理学院,陕西咸阳712000

出　　处：《微型电脑应用》2021年第5期52-54,68,共4页Microcomputer Applications

基　　金：陕西青年自然科学基金项目(18SX091)。

摘　　要：针对现有股票价格预测方法存在预测精度低和泛化能力差等缺点,提出了一种CFOA优化FSVM的股票价格预测方法。通过CFOA对FSVM模型的参数Gamma、b自适应寻优,从而实现股票价格的预测研究。为验证算法的有效性,选择2010年3月30日—2020年3月20日的上证综合指数为研究对象,采用滚动预测的方式进行上证综合指数预测研究。实验结果表明,以预测时间和预测均方误差为评价指标,CFOA-FSVM比FSVM和FSVM具有更高的预测精度,从而验证了CFOA-FSVM进行股票价格预测的有效性和可靠性。研究结果可为及时和实时掌握股市行情,提供科学决策的依据。Aiming at the shortcomings of the existing stock price forecasting methods,such as low precision and poor generalization ability,a CFOA optimized FSVM stock price forecasting method is proposed.Through CFOA adaptive optimization of parameters Gamma,b of FSVM model,the prediction of stock price is realized.In order to verify the algorithm,the Shanghai Composite Index on March 30,2010 and March 20,2020 was chosen as the research object,and a rolling forecast method was used to forecast the Shanghai Composite Index.The experimental results show that CFOA-FSVM has higher prediction accuracy than FSVM and FSVM,which proves the validity and reliability of CFOA-FSVM in stock price forecasting.The research results can provide the scientific decision-making basis for the timely and real-time grasp of the stock market.

关 键 词：果蝇优化算法 混沌理论 模糊支持向量机 股票价格 预测精度 评价指标 

分 类 号：F830.91[经济管理—金融学]