Research On Stock Allocation Strategy Based On Clustering And Decision Tree Algorithm

With the development of China's market economy and the continuous deepening reform of various systems,China's capital market is gradually developing and improving.The stock market has become an important part of China's capital market.People's financial and investment awareness are increasing.Investment has become an important way for people to allocate assets.Therefore,how to choose stocks,balance risk and return,and rationally optimize stock allocation has very significant theoretical significance and very attractive application value.The stock allocation refers to the allocation of investment funds among different stock categories according to investment demand.The essence is to analyze the stocks and then select one or a group of stocks according to their risk tolerance and investment style to achieve reasonable expectations of investment risks and returns.However,the securities market itself is an extremely complicated system,including high noise,non-linearity and blindness of investors.These factors determine the complexity of stock price forecasting.The problem of stock allocation decision-making is always social and academic.One of the hot research areas,so how to rationally choose stock allocation is also a hot issue that has not been solved so far.This paper mainly constructs the model based on data mining algorithm and valueinvestment theory.Through the analysis and comparison of the model results,the paper will evaluate the effect and usability of the model.Finally,combined with the analysis results of the data model,the paper will provide stock allocation strategy andinvestment opinions.On the basis of value investment,the article studies the fluctuation of stock price and the strong correlation between certain indicators of listed companystocks.In the specific research process,the sample data selected in this paper is mainly the Shanghai-Shenzhen 300 stock that is more suitable for the theoretical analysis ofvalue investment.In the selection of data mining algorithm model,this paper mainlychooses cluster analysis and decision tree algorithm to construct the model.Clusteranalysis can describe the category features very well.Therefore,the cluster model can describe the characteristics of stocks with different profit performance and providedecision direction for stock allocation.The decision tree algorithm can well performdata nodes based on data features.Classification,and can also generate classificationrules to make classification predictions for individual data nodes,so the decision tree model can intuitively predict stock earnings performance.The above two algorithms perform stock analysis from the macro level and the micro level respectively,and the two are combined to complete the comprehensive research and analysis of this paper.