基于技术指标和随机森林的股价走势预测算法

Stock Price Trend Prediction Algorithm Based on Technical Index and Random Forest

由于股票价格波动的复杂性和动态性,预测股价走势多年来一直是研究人员关心的领域。将预测问题视为分类问题,以股票的异同移动平均线、平均趋向、相对强弱、布林线、强力指数五个技术指标和下周股价走势作为随机森林预测模型的特征,然后通过网格搜索优化随机森林模型的参数,构建基于技术指标的GS-RF股价走势预测模型。实验结果表明,相比于技术指标交易策略的收益率,使用GS-RF模型收益率最高、风险最小。

Due to the complexity and dynamics of stock price fluctuations,predicting stock price movements has been an area of concern for researchers for many years.Regarding the prediction problem as a classification problem,MACD,ADX,RSI,BB,FI and next week move are taken as the characteristics of the random forest prediction model.Grid search optimizes the parameters,thus constructing a GS-RF stock price trend prediction model based on technical indicators.The experimental results show that compared with the return rate of technical indicator trading strategies,using the GS-RF model has the highest return rate and the least risk.