基于集成学习的高送转股票研究

发布时间：2018-01-20 16:20

本文关键词： 高送转集成学习非平衡数据投资组合　出处：《时代金融》2016年36期 　论文类型：期刊论文

【摘要】：高送转预案公告发布前,高送转股票具有显著的累计正收益,因此预测高送转股票对于投资具有重要意义。高送转股票的预测是分类预测问题,本文利用上市公司三季度财报数据,采用3种集成学习算法:由K-近邻算法、决策树以及加lasso惩罚项的逻辑斯蒂回归算法构建预测模型——"组合"模型,经典的集成学习算法——Ada Boost算法以及随机森林算法进行建模。本文采用准确率以及G-mean作为模型评价标准,结果显示:"组合"模型的准确率最高,随机森林和"组合"模型的G-mean表现相当,均优于adaboost算法。由于每年高送转股票所占比例小于50%,数据可以看成是非平衡数据,为了改善"组合"模型较差的召回率,本文采用K-Means聚类的欠抽样方法,将此方法用在"组合"模型上,效果显著。最后分别对上面三种模型预测的股票构建投资组合,并以HS300指数做基准。结果显示:"组合"模型预测得到的高送转股票组合表现优于另外两种集成学习模型。
[Abstract]:Before the announcement of the high transmission plan announcement, the high transmission stock has significant accumulative positive income, so it is important to forecast the high transmission stock for investment. The forecast of the high transmission stock is the problem of classification forecast. In this paper, we use three integrated learning algorithms based on the third-quarter data of listed companies: the K-nearest neighbor algorithm. The decision tree and the logistic regression algorithm with lasso penalty term are used to construct the prediction model-" combination "model. The classical ensemble learning algorithm, Ada Boost algorithm and random forest algorithm, are modeled. In this paper, the accuracy and G-mean are used as the model evaluation criteria. The results showed that the accuracy of the "combination" model was the highest, and the G-mean performance of the random forest model and the "combined" model was the same. The data can be regarded as unbalanced data, in order to improve the poor recall rate of the "combination" model. In this paper, the K-Means clustering method of under-sampling is used in the "combination" model, and the effect is remarkable. Finally, the portfolio of stocks predicted by the above three models is constructed. The results show that the performance of the "portfolio" model is better than the other two integrated learning models.
【作者单位】：华南理工大学数学学院;
【分类号】：F832.51
【正文快照】： 一、引言所谓“高送转股票”是指上市公司大比例送红股或大比例以资本公积金转增股票,市场送转股比例超过0.5的股票为“高送转股票”。虽然上市公司送股、转增股票及不影响其当期现金流,也不影响其未来现金流,从而这种分红并不影响公司价值,但高送转事件向市场传递了公司发

【相似文献】