Cluster Analysis Of Data For The Securities Industry And Its Application

Cluster analysis is an important method of multivariate statistical analysis.Clustering techniques can be used to explore the data in hidden structural features.On stock market,the efficiency of information acquisition and utilization is directly related to investment return.This paper uses cluster analysis to discover the internal structure of stock data and visualize the results.(1)Clustering analysis using a variety of features to measure stock data.With cluster analysis,we explore the underlying patterns of different stocks,which will help investors to establish investment scope and ultimately realize investment value.(2)A variety of cluster analysis techniques have been introduced and compared in this paper.We found the hierarchy clustering method is more intuitive and has better visualization.(3)Time series data is another very important stock market data.This paper also discussed a solution to clustering daily close price data.The paper illustrate a simple use case of time series clustering by clustering time series of SSE50 stocks and GEM stocksBoth theoretical and experimental results proved the importance of stock market data clustering analysis.With the internal patterns found out by clustering analysis,investors can better understand the market and potentially make better investment decision.