The Research Of State Transition Model Based On The Stock Market Volume And Price Distribution

In the context of computer technology’s developing and big data, methods ofmathematics, physics, and data mining are being wildly used in economics andfinance. They have partly solved two problems in the past researches: the limits oftraditional economics and finance qualitative and quantitative research methods; andthe shortcomings of hypothesis of rational man and efficient market which can’taccount for the reality. They have pointed out for the researches of economics andfinance a total new prospective and directions.In this paper, mathematical methods and data mining techniques are used toexplore the approximate normal form of Chinese stock market trading volume andprice distributions. Chinese stock market trading approximate normal distributions involume and price distribution for different time spans are studied, and on the basis ofthese two studies state transitions model for trading volume and price distribution isproposed.The main works of this paper consist of three parts: First, the hypothesis that theChinese stock market day trading volume and price distributions refuse to obeynormal distribution is verified, noting that the reason of the refusal is the unusuallarge volume in some prices, however, a "fat tail" approximate normal distributionpattern is founded. Then, after the smoothing the distribution of volume and price,approximate normal distributions of trading volume and price distributions fordifferent time spans are founded. Finally, on the basis of two previous studies, statetransition model of the stock market volume and price distributions is proposed.Based on the state transition probability table that is created by clustering analysis, thefuture approximate normal distributions structure that may occur is predicted.Transition model of the state of the stock market volume and price distributionsestablished in this paper can appropriately account for the transition Chinese stockmarket trading volume and price distribution network transitions between states, and can be used in a certain probability to predict possible future distribution of volumeand price, to provide a quantitative indicator in the stock market investment.