Stock Index Price Trend Prediction And Investment Strategy Analysis Based On Support Vector Machine

In recent years,people decide the investment and trading behavior by means of quantification,and choose the optimal strategy for trading by simulating and quantifying the combination of different stocks and trading time.Therefore,the analysis of stocks and how to accurately predict the stock price have become the key to making investment decisions.The change of stock price is a non-linear change process.Traditional financial prediction methods,such as fundamental analysis and time series analysis of technical analysis,have some limitations,and cannot describe the change process of stock price well and make corresponding prediction.Based on this,the support vector machine(SVM)model is used to model stock price changes and carry out corresponding strategy analysis,so as to give full play to its advantages in small samples and nonlinearity.In order to mine more information in the data,the Shanghai composite index was used in this paper.Firstly,outliers were identified and screened from the sample data,and different support vector machine models were established using genetic algorithm and particle swarm optimization algorithm.Then,corresponding investment strategies were formulated through simulation experiments.Setting the control group by using GARCH model to contrast with SVM model data and results.The results show that the model of support vector machine based on particle swarm optimization algorithm can predict the trend of stock price more accurately.At the same time,outlier recognition can mine more data information and improve the prediction efficiency of the model.