机器学习算法在股市股票收益率变化预测中的应用研究

Research on the Application of Machine Learning Algorithm in the Prediction of Stock Return

针对当前基于机器学习算法的股市股票收益率预测方法精度和效率不够理想的缺陷,提出了一种股市股票收益率预测模型。首先,针对股票数据中含有较多噪声的问题,采用小波变换算法对其进行降噪处理。然后,基于已有的研究成果和股市现状,构建了股票收益率预测指标体系。最后,针对SVM的缺陷,采用一种改进的果蝇算法对其进行优化,并构建股票收益率预测模型。测试结果显示,IFOA-SVM模型的误差值为0.18,Loss值为0.23,预测精度达到95%,AUC值为0.954。研究提出的方法能够准确预测股票收益率,从而提高投资者收益,刺激股票市场的发展。

In view of the shortcomings of the accuracy and efficiency of the current stock return prediction method based on machine learning algorithm,a stock return prediction model is proposed.First of all,in view of the problem that the stock data contains more noise,the wavelet transform algorithm is used to denoise it.Then,based on the existing research results and the current situation of the stock market,the stock return prediction index system is constructed.Finally,aiming at the defects of SVM,an improved Drosophila algorithm is used to optimize it,and a stock yield prediction model is constructed.The test results show that the error value of IFOA-SVM model is 0.18,the loss value is 0.23,the prediction accuracy is 95%,and the AUC value is 0.954.the method proposed in the study can accurately predict the stock return rate,thus improving investor returns and stimulating the development of the stock market.