Research On The Application Of Data Mining Technology In Stock Analysis

The stock market,as the important part of the market economy,has been the main way of investing financial management.How to choose high-quality stocks and the best investment point to maximize profits is concerned by investors.With the development of information age,the stock market produces massive data every day.The traditional stock analysis methods can no longer meet the requirements.Data mining technology provides an effective way for the stock analysis.The thesis is divided into three parts: Firstly,the basic contents of data mining technology and stock analysis are introduced,and the status of data mining technology applied to stock analysis is summarized.Secondly,the thesis introduces the algorithms used in data mining technology,including association rule algorithm,decision tree algorithm and neural network algorithm.Finally,the above algorithms are used to analyze the stock data,which is the important part of the thesis.The main contents of the empirical analysis in the paper is choosing 2301 valid stocks of all A stock market in 2016 for analysis.firstly,using association rule analysis to choose 17 financial indicators,which is more related to the increasing of stock,from the 24 financial indicators that affect stock;Secondly,according to the key indicators selected,the 2301 stock is classified by the decision tree algorithm,and the most valued 231 stocks are obtained,which can provide reference for investors in choosing stock;Finally,the thesis using neural networks methods to predict the trend of stocks' price.The thesis randomly selected some good stocks to carry out experiments.Using BP neural network method,RBP neural network method and RBP-BP combined neural network method to predict the trend of stocks,and the experimental results show that RBP neural network behaves worst while RBP-BP combined neural network gives the best prediction,which can provide investors a reference about the best investment point.Experimental results show that using data mining technology to analyze the stock market can overcome the shortcomings of traditional data analysis methods to certain extent and provide the financial management reference for investors.