基于集成算法的股票指数预测
    点此下载全文
引用本文:王 玥,孙德山.基于集成算法的股票指数预测[J].经济数学,2018,(4):28-30
摘要点击次数: 1230
全文下载次数: 0
作者单位
王 玥,孙德山 (辽宁师范大学 数学学院 辽宁 大连 116029) 
中文摘要:利用集成算法中的Bagging、Boosting和Random Forest三个方法,选取股票指数中的中小板指数、深证成指数、上证指数、创业板指数4组数据进行分析,得出Random Forest对上证指数、中小板指预测结果较好;Boosting对创业板指预测结果较好;Bagging对深证成指预测较好.并在4个板指中,随机选取了4支股票数据(分别为大连重工、中南建设、中国医药、东方国信)进行分析,得出集成算法在数据为200个的情况下,预测结果较为准确,其中不同方法对不同股票的适宜程度有所不同.
中文关键词:股票指数  袋装法  提升算法  随机森林
 
Stock Index Prediction Based on Ensemble Algorithm
Abstract:Using the three methods of bagging, boosting and random forest in the integrated algorithm,this paper selected the small and medium-sized board index, the Shenzhen Stock Exchange Index, the Shanghai Composite Index, and the GEM Index for analysis. It is concluded that random forest has better prediction results for the Shanghai Composite Index and the small and medium board index; boosting has better prediction results for the GEM index; bagging has better predictions for the Shenzhen Stock Exchange Index. And among the four board fingers, four stock data were randomly selected (Dalian Heavy Industry, Zhongnan Construction, China Medicine, Dongfang Guoxin) for analysis. It is concluded that the integrated algorithm has more accurate prediction results when the data is 200. The suitability of different methods for different stocks is different.
keywords:stock index  Bagging  Boosting  Random Forest
查看全文   查看/发表评论   下载pdf阅读器