Research On Intelligent Securities Recommendation Method Based On User Investment Style

As the investor group and the financial market information scale continue to expand,the traditional Internet financial platform with financial data as the core has been unable to satisfy the information requirements of various investors,making the research core of the Internet financial platform gradually change from providing financial data information to providing financial services.Securities recommendation as an investment suggestion service can help investors make better investment decisions,which makes it play an increasingly important role in the field of securities investment.Current securities recommendation system has a series of problems such as serious system resource consumption,difficulty in real-time online securities recommendation,poorly recommended securities,and low quality of recommended securities.In order to solve the above shortcomings,this paper uses the audience of the securities recommendation system----Investors come to the main research object for investor style portrait modeling,investor group relationship mining and intelligent securities recommendation modeling,and carry out intelligent securities recommendation research based on user investment style.In view of the limitation that the current securities recommendation system can not accurately obtain the important characteristics of investors,this paper proposes an investment style portrait modeling method based on user investment analysis,which combines investor investment data and market data with traditional performance attribution theory.The investment ability and investment preference are used to establish a more complete description of the investor style.The corresponding investor style portraits are designed according to the characterization,and the timeliness indicators of the features are established to verify the stability of the portrait description,the effectiveness and feasibility of the system implementation and platform application verification modeling method.The traditional securities recommendation system can't dynamically capture the user's interest offset status,which seriously affects the adaptability of securities recommendation results to investors.This paper uses the style characteristics of investors and adopts the method of group relationship mining to dynamically capture the investor's investment style.In the case of transfer,an optimized genetic clustering group mining model is proposed.The traditional group mining clustering model is deeply investigated,and the genetic components of the basic genetic clustering algorithm are optimized,including chromosome initialization optimization,fitness function design optimization,genetic operation optimization and other optimization strategies.Based on this,the investment is designed.The overall scheme of group division,including system module composition,operation logic and so on.According to the optimized design scheme,the related experiments are carried out,which proves that the proposed model can show better performance than the traditional clustering mining method.After completing the investor description and capturing the investor characteristic offset,this paper proposes a securities recommendation model based on investor investment style and group relationship.The recommendation algorithm based on singular value decomposition theory is deeply studied.Aiming at the defects of high computational complexity and poor scalability,a recommendation method combining user group division is proposed.The technique of matrix dimensionality reduction and investor group division is adopted to reduce the cost of the online recommendation algorithm for system cost.By inrtoducing the two-layer SVD mechanism and the investor preference capture mechanism to improve the recommendation effect of the securities recommendation model.The proposed accuracy recommendation model and the recommended securities revenue performance indicators verify that the proposed securities recommendation model can provide investors with high-quality,personalized online securities recommendation services.