基于财务指标的股价涨跌预测模型  

Stock price prediction model based on the financial index

作　　者：刘新月 程希明[1] LIU Xinyue;CHENG Ximing(School of Applied Science,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学理学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2022年第1期96-100,共5页Journal of Beijing Information Science and Technology University

摘　　要：针对创业板短期股票价格涨跌趋势的预测问题,提出了利用财务指标进行分析的股价涨跌预测模型。通过分析股票市场的影响因素,确定了由6类财务指标构建的特征变量体系;为消除特征变量之间的共线性,对特征变量应用递归特征消除法、主成分分析法等进行降维;改进测试集选取方法,选用与训练集具有相同季节特征的数据作为测试集,以消除季节性影响;选取随机森林、支持向量机、岭回归3种机器学习算法建立数学模型进行对比分析。结果显示,主成分分析与支持向量机的组合模型表现最好,其评价指标中测试集精确度为0.771 812,表现较好;其KS指标在训练集和测试集上的差值为0.002 45,低于其他模型,表明该模型对创业板股票数据的适配性最高。Aiming at the prediction of short-term trend of stock price rise and fall on GEM,a prediction model based on the financial index is proposed.By analyzing the influencing factors of the stock market, the characteristic variable system constructed by six types of financial indicators is determined.In order to eliminate the collinearity between characteristic variables, the recursive feature elimination method and principal component analysis method are used to reduce the dimension of characteristic variables.The test set selection method of the machine learning algorithm is improved, and the data with the same seasonal characteristics as the training set is selected to replace the traditional algorithm to separate part of the data from the training set as the test set, so as to eliminate the seasonal influence.Three machine learning algorithms including random forest, support vector machine, and ridge regression are selected to establish mathematical models for comparative analysis.The results show that the combined model of principal component analysis and support vector machine performs best, and the accuracy of the test set in its evaluation index is 0.771 812.The difference of KS index between the training set and test set is 0.002 45,which is lower than that of other models, indicating that the model has the highest adaptability to gem stock data.

关 键 词：数据挖掘 财务指标 测试集划分 机器学习算法 股价预测 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] F830.91[自动化与计算机技术—控制科学与工程]