摘要
文章通过比较分析价值策略和成长策略,提出了以价值成长投资策略(GARP)理念为基础的选股模型指标体系,选用了2012年1月至2013年2月间360多支股票的4406个样本数据,通过等频算法对数据进行离散化预处理后,使用随机森林算法实现了较高正确率的股票分类,投资者可以据此判断是否继续持有股票。通过分析优选后的股票在行业平均收益、最值方面的实际表现,验证了该量化选股模型性能优异。
By comparing the analytical value strategies and growth strategies, the paper proposed the indicator system of the stock selection model based on the Growth at a Reasonable Price （GARP） and selected four thousand four hundred and six samples＇ data of more than 360 stocks from January 2012 to February 2013. After the discretized preconditioning for the data through the algorithm of equivalent frequency, the paper achieved a higher accuracy of stock classification by the random forest algorithm. Investors can judge whether to continue to hold the stock. The paper validated the performance of the stock selection model by analyzing the actual performance values in the average income, the minimum and maximum values in the industry.
出处
《首都经济贸易大学学报》
北大核心
2014年第2期21-27,共7页
Journal of Capital University of Economics and Business
基金
国家自然科学基金面上项目<基于预测建模的宏观经济时间序列结构变化研究>(项目编号71071022)
关键词
随机森林
股票选择
股票投资
价值成长投资策略
random forest
stock options
equity investments
value growth investment strategy
