基于信息粒化和BP-Bagging的股价范围预测模型  

Prediction of stock price range based on information granulation and BP-Bagging

作　　者：赵雅博 何华[1] ZHAO Yabo;HE Hua(School of Science,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学理学院,天津300401

出　　处：《河北工业大学学报》2022年第5期55-59,74,共6页Journal of Hebei University of Technology

基　　金：国家自然科学基金(11171087)。

摘　　要：针对股票价格点预测的复杂性和不确定性,以及单BP神经网络易出现过拟合的问题,提出了一种基于模糊信息粒化和BP-Bagging的股票价格波动范围的集成预测模型。首先对训练数据进行模糊信息粒化处理,根据预测模型的需要提取各个窗口的模糊信息。然后随机有放回的抽取若干个子训练集,用于训练不同的BP神经网络模型。最后结合Bagging算法对多个网络进行集成生成强预测器,提高预测模型的准确性。在此基础上对股票开盘价的波动范围进行预测,将此模型与单BP神经网络与支持向量机(SVM)的方法进行比较,实验结果表明,基于BP-Bagging的强预测模型具有较好的稳定性和时效性。In view of the complexity and uncertainty of stock price point prediction and the over fitting problem of single BP neural network,an integrated prediction model of stock price fluctuation range based on fuzzy information granulation and BP Bagging is proposed.First,the training data is dealt with fuzzy information granuation,and the fuzzy information of each window is extracted according to the needs of the prediction model.Then,several sub training sets are randomly extracted with retrieval to train different BP neural network models.Finally,the Bagging algorithm is used to integrate multiple networks to generate a strong predictor to improve the accuracy of the prediction model.Based on the above,we predict the fluctuation range of the stock opening price,and compare this model with the methods of single BP neural network and support vector machine,The experimental results show that the strong prediction model based on BP-Bagging has good stability and timeliness.

关 键 词：模糊信息粒化 BP-Bagging SVM 股票价格 范围预测 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] F830.9[自动化与计算机技术—控制科学与工程]