Research on Comprehensive Analysis Method of Stock KDJ Index based on K-means Clustering

Abstract

This paper proposed a K-means measure to cluster stocks, and predicted the investment of strong profitability objects through comprehensive analysis of KDJ indicators. The paper analyzed the clustering hierarchy diagram, as well as the inter-cluster similarity structure diagram of different cluster numbers. It is found that the clusters can be effectively distinguished for each type of stock. The comprehensive prediction precision of KDJ are better than each single index. The feasibility and effectiveness of the suggested method are verified by the example of the constituents of the CSI 800 Index. The quantitative investment model established by the analytical method in this paper has better prediction effect.