Research On Intelligent Evaluation Of User Investment Behavior And Portfolio Selection Method

With the continuous development of our society and economy,the continuous improvement of the people's living standard and the continuous improvement of the financial market system in our country,more and more people are participating in the securities investment.According to a survey,120 million people in China have been involved in securities investment by the end of 2016.At the same time,Chinese investors not only have the characteristics of being regional,younger and less educated,but also have strong blindness and herd mentality.Although the stock market of China is growing,the overall living conditions of Chinese investors are not good.With the continuous development of computer technology,computers have changed people's lifestyles in many fields and greatly enhanced the production and living efficiency.However,the development of platform systems for the securities investment evaluation is still relatively scarce.Currently,only the brokerage platform and some mature securities market analysis software come with simple user investment behavior analysis functions,and the analysis of user investment they provide is limited to the calculation of some statistical indicators and the general summary of investment features.In general,the securities investors are most concerned about two issues,benefits and risks.On the one hand,it studies how to carry out scientific,reasonable,systematic and comprehensive investment behavior evaluation to securities investors;on the other hand,it studies how to use intelligent optimization algorithm to provide high-yield,low-risk portfolio solutions.Aiming at investment behavior evaluation,this paper presents a quantitative evaluation method of user investment behavior.By processing the investment records of users,a vectorized representation of the user's investment interest is obtained.Attempt to use k-means algorithm,DBSCAN algorithm to cluster user interest.After cluster analysis of user data from many sources such as internet crawling and simulation investment platform,it is found that securities investors show the similarities of groups in investment habits and interests.At the same time,it is found that the DBSCAN algorithm has a better clustering effect when using the contour coefficient as the evaluation index.According to some basic principles in the field of securities investmen,with reference to some methods and models of quantitative investment,six quantitative indicators are selected as the evaluation indicators for the investment behavior of users.And according to the six evaluation index value,gives the user investment behavior comprehensive ability evaluation radar chart.Through the design and implementation of user investment behavior evaluation platform,the feasibility of quantitative evaluation method of user investment behavior is verified.Aiming at portfolio optimization,the subject takes the mean-variance portfolio model as the theoretical basis and takes the two intelligent algorithms,whick are particle swarm optimization and genetic algorithm,as the optimization methods to carry out the experiment and analysis of portfolio optimization.Particle swarm optimization algorithm and genetic algorithm are used to search the optimal portfolio of nine stocks in CSI 300.Through a large number of comparative experiments,it is found that PSO has better effect and faster convergence rate in portfolio optimization.