The Research Of Stock Data Dynamic Evolving Network

In the study of the stock market,through the economic theory analysis of scholars,we can conclude that the stock market changes with the macro-factors,micro-factors and industry factors.Therefore,when using complex networks to analyze the stock market,it is necessary to consider the overall trend of the market in different economic environments.So,we need to use the dynamic evolution of the network research methods under the influence of different economic factors in the market research and analysis.At present,the research on dynamic evolution network is still in the exploratory stage,which mainly analyzes the topological attributes at the network level.However,the dynamic evolution process is a collection of changes in the connection between nodes in a network.Therefore,it is impossible to analyze and analyze nodes in a dynamically evolving network from a single network.Therefore,this paper proposes to combine dynamic evolution network with clustering analysis method,and use clustering method to cluster analysis of the change of nodes between networks.This article uses the transaction data from 2013 to 2015 in the Chinese stock market and deals with 924 stocks with high-frequency trading properties.Pearson correlation coefficient is used to define the correlation between stocks and construct the network.Innovatively,the method of hypothesis testing is used to complete the network simplification,and the complex network of Chinese stock market is successfully constructed in six cycles.By comparing and analyzing the topological attributes of each periodic network,we can draw some rules:First,when the market environment is more turbulent,the correlation between stocks tends to be more positive and the network more central;Second,the performance is more obvious when the overall trend is declining during the turbulent market environment.Third,before the turbulent turbulence in the stock market,the stock network is more divergent.And further study using clustering analysis of the stock nodes,we can find that the Chinese stock market can be divided into three main categories of stocks,namely,stable stocks,rising market-sensitive stocks and market-sensitive stocks.