混沌映射同步的上市公司聚类分析

Clustering analysis of listed companies in stock market based on chaotic map synchronization

对股票市场中众多的上市公司进行识别聚类分析,有助于制定出较优的投资组合策略.采用混沌映射聚类算法,根据上市公司的股票价格建立相关映射,并且将该金融时间序列的相关系数与映射之间的耦合强度联系在一起进行分析.在以香港恒生中国内地25指数成分中的上市公司为样本的实证研究中,为了通过股票交易价格识别出上市公司的相似状态,运用了两两的聚类分析比较方法.研究结果表明,对混沌映射动力学的模拟可以使数据自然的分割,属于相同产业背景下的公司通常是聚类在一起的.

Identification and clustering analysis for the listed companies in stock market is helpful to set down portfolio optimization strategies.In this paper,chaotic map clustering algorithm was used to establish correlated maps based on the stock prices of the listed companies,and the coupling strength associated with maps was related to the correlation coefficients of the financial time series.A pairwise clustering approach is applied to identify similarity of the listed companies which constituted Hang Seng mainland 25 index by using their stock prices.This empirical study shows that the simulation of a chaotic map dynamics leads to a natural partition of the data,since companies involving in the same industrial branch are usually grouped together.