Research On Self-Learning Preferences Model Of Interactive Decision In Portfolio Problem

During the time of the rapid economic development and the continuous improvement of the quality of people’s life,the education level of people has also been continuously improved,and more and more people are focusing on how to invest in financial management.Therefore,how to scientifically manage money matters and put limited capital into financial products to obtain higher returns with lower risks is a problem that every investor is thinking and exploring.Investment portfolio can not only improve the income,but also can control the risk to a certain extent.Therefore,the investment portfolio is the core issue in the investment field.The essence of the investment portfolio problem is a multi-objective optimization problem.In the existing research on its solutions,most of them can only obtain a group of effective frontiers with non-inferiority,but as a result of the decision makers cannot rationally choose the most suitable one for their own from them.So the solution to the problem is that the existing research methods cannot fundamentally solve the issue of program selection when investors make investme nt decisions.In order to better assist decision makers in obtaining the best investment plans fitting for their own preferences,considering the limited rationality of decision makers,this paper proposes an interactive multi-criteria decision-making method.This method includes three main links: multi-objective evolution method,preference feedback and preference model.Among them,the preference model link is used to learn the preference information of decision makers obtained from the preference feedback,and guide the follow-up solution in the form of decision-oriented guidance,so that interactive multi-criterion decision-making results can be obtained by a few or even the only one meeting the needs of the decision makers’ personal preference decision plan.Therefore,the authenticity,flexibility,and orientation of the preference model play a crucial role in the effectiveness and efficiency of interactive multi-criteria decision-making.However,the current preference model mostly adopts the form of a function or neural network,which has a very high quantitative requirement for the preference information provided by the decision maker,and the fixed function form limits the flexibility of the preference model.Therefore,based on the existing research on multi-objective evolutionary methods and preference feedback,this paper makes an in-depth study of the preference model.A preference model in the form of a decision tree was established,self-learning was used to update it,and a decision-oriented mechanism of the preference model was used to guide the interactive multi-criteria decision-making process.Based on the study of preference model theory,the interactive multi-quasi-decision making method based on preference model is applied to the actual investment portfolio problem,and its validity test is carried out.Through the validity test of the preference model,the authenticity,flexibility and guidance of the self-learning preference model are proved,and the interactive multi-criteria decision-making based on the preference model can be used to help the decision makers find the only investment option for their own that is preferred.