Research On Stock Data Analysis Based On Data Mining Technology

Stock market has undergone nearly 30 years of development in china and has become an important part of China's capital market.At the same time,with the improvement of people's income level,stock investment has become an important means for investors to allocate assets.The complex and volatile stock market contains a lot of very huge data.How to make better use of this data and mine the deep information in the data has become a new challenge.For a long time,it has been difficult for ordinary technical means to analyze investment methods to mine the information hidden behind huge amounts of data.Using data mining technology to analyze stocks can better mine potential useful information and value laws of listed companies' stocks,and help investors make correct investment decisions.According to the characteristics of stock data,this paper selects the corresponding data mining model for stock analysis and prediction,analyzes and solves the problem of what stocks investors choose to have higher returns.It is mainly divided into three parts:first is to find the key indicators.Because there are many factors that affect the stock price trend,This article uses association rule analysis to find 20 key stock indexes that have reference significance and affect stock price trends.Improved effectiveness,avoiding too many indicators affecting analysis results.Then,based on the results of the association analysis,these key indicators are clustered.By comparing different clustering algorithms,the K-Means algorithm is selected for clustering and algorithm improvements are made.The accuracy of the improved K-Means algorithm is significantly improved.According to the results of cluster analysis,it can be seen that high-quality stocks are better clustered,distinguishing high-quality stocks from inferior stocks,and playing a more effective role in guiding investors' investment.Finally,a decision tree algorithm is used for predictive analysis.The decision tree algorithm is classified according to the characteristics of the data.At the same time,a single data node can be predicted.Therefore,the decision tree model can be used to predict the stock trend intuitively.After experimental analysis,the improved C5.0 decision tree can effectively improve the accuracy of the prediction through the C5.0 decision tree combined with the Boosting algorithm,and the addition of two input indicators is also effective for improving the accuracy,which is applied to the prediction of individualstocks.It has a certain guiding significance.This paper uses data mining techniques to analyze stock data from the macro and micro levels by using different algorithms.It solve the problem of traditional stock analysis methods to a certain extent and provides a valuable reference method for stock investment forecasting.