Application of Parallel Clustering Algorithms for Big Data in the Division of Stock

Abstract

For the operating performance of listed corporations reflects the value of stock investment to a certain extent, financial index reflecting the operating performance of listed corporations was taken as the evaluation index of stock investment value, and for the first time the parallel clustering algorithms for big data both K-means and fuzzy K-means of Mahout were used to cluster nearly 2 600 stock of China’s A shares market according to their financial index, afterwards the clustering results of these two algorithms under different distance metrics were compared.Experimental results show that the clustering quality of K-means algorithm adopting Tanimoto distance metric is the best.Therefore, this result can be used as the final result of the division of stock, which can provide a reference for the investment decision.

Keywords

• inancial index;parallel clustering algorithm;K-means;fuzzy K-means;division of stock