摘要
研究了一种附有引力影响因子的半监督K-means核函数聚类算法,并将该方法应用于多因子选股模型中。研究表明,相比传统的聚类模型,改进的模型具有较强的泛化能力,模型在处理样本线性不可分、样本分布非球状簇等问题上具有明显的优势,能选出较优的股票组合。
This paper studies a kind of semi-supervised K-means kernel function clustering algorithm with gravitational influence factors,and applies it to the multi-factor stock selection model.The empirical evidence shows that:Compared with the traditional clustering model,this improved model has a strong generalization ability,and it has obvious advantages in dealing with the non-linear and non-spherical of sample distribution,and can select a better stock combination.
出处
《统计与信息论坛》
CSSCI
北大核心
2018年第3期30-36,共7页
Journal of Statistics and Information