Improved v -Support vector regression model based on variable selection and brain storm optimization for stock price forecasting

Display Omitted Original v-support vector regression is improved from two aspects.Principal component analysis is used to select suitable inputs for v-SVR.Brain storm optimization is first used to optimize three parameters of v-SVR.Two case studies of representative Chinese stock indices are presented.The developed hybrid model outperforms competing models for stock price forecast. Big data mining, analysis and forecasting always play a vital role in modern economic and industrial fields, and selecting an optimization model to improve time series forecasting accuracy is challenging. A support vector regression (SVR) model is widely used forecasting and data processing, but the individual SVR cannot always satisfy the requirements of time series forecasting. In this paper, a hybrid v-SVR model is developed and combined with principal component analysis (PCA) and brain storm optimization (BSO) for stock price index forecasting. Correlation analysis and PCA are conducted initially to select the input variables of the v-SVR from 20 technical indicators, while the advanced BSO algorithm is used to search for optimal parameters of v-SVR. Case studies of the China Securities Index 300 (CSI300) and the Shenzhen Stock Exchange Component Index (SZSE Component Index) are examined as illustrative examples to evaluate the effectiveness and efficiency of the developed hybrid forecast strategy. Numerical results indicate that the developed hybrid model is not only simple but also able to satisfactorily approximate the actual CSI300stock price index, and it can be an effective tool in stock market mining and analysis.