基于GA-SVM的股票价格预测分析

Stock price is nonlinear, stochastic, complex and other characteristics, so that the traditional stock prediction methods such as statistical methods, linear regression, time series and other prediction methods are no longer suitable for stock price prediction. In order to quickly and accurately analyze and forecast stock prices, this paper adopts 598 groups of basic stock price data of Ping An Stock from March 14, 2018 to August 27, 2020.In order to reduce the training complexity, principal component analysis (PCA) is used to reduce the dimension of the original data, and genetic algorithm (GA) is used to support vector machine parameter optimization (SVM). After predicting the closing price of 598 groups of Ping An stocks, it was found that the mean square error of the test group obtained by USING PCA-GA-SVM was 0.0032 and the determination coefficient was 0.98814, which made the stock price prediction fast and accurate.