Based On The Ant Clustering Class Stock Plate Classification Research Of The Algorithm

As the China stock market developing, it is more and more important to classify stocks in order to construct portfolio for reducing risk.According to the modern portfolio theory, people can reduce the idiosyncratic risk by constructing a portfolio. The risk of the portfolio is related to the correlation between portfolio equities, the smaller the correlation between the stocks the more obvious dispersion effects portfolio has. Now, it is normally to classify stocks by industry in China. So, if the correlation between stock returns within the same industry is higher than the correlation between stock returns of different industries, portfolio built with stocks from different industries should be less risky.However, this paper has proved that stock price volatility has a high correlation between industries in China stock market through empirical research. So, it is less effective to build a portfolio by industry. Therefore, it is very important to get a new stock classification standard other than industry.This paper used the optimized ant colony clustering algorithm to perform cluster analysis on the more than2,000stocks of China’s A share market both with financial ratios and stock price volatility as the classification criteria, and got the conclusion that the clustering result is efficient. So, this paper has provided a new way to classify stocks efficiently and to make investment decision and risk controlling.This article consists of five chapters:Chapter one is Background, literature review and the research framework; The second chapter describes stock classification by industry in the Chinese stock market and its problems; Chapter three is about the classification based on the optimized ant colony clustering algorithm; Chapter four gives the empirical research result of the method given by chapter three; chpater five is the conclusions and outlooks.