Research On Fuzzy Portfolio Model Based On Intelligent Algorithm And Its Application

As an extremely complex system,the securities market has a lot of uncertainty,and uncertainty is the difficulty in decision analysis research.There are two main forms of uncertainty in events: randomness and ambiguity.Portfolio research based on stochastic uncertainty has been developed quite well,and portfolio research based on fuzzy uncertainty has gradually been recognized and paid attention to.This paper will use fuzzy set theory,optimization method and intelligent optimization method to study the fuzzy portfolio selection problem.The main research work and innovations are summarized as follows:(1)A variety of effective intelligent algorithms or numerical methods are designed for the model being constructed.(1)Proposed a new PSO-AFSA hybrid algorithm.The results of the particle swarm algorithm search are used as the initial fish population of the artificial fish swarm algorithm,and the best position in the artificial fish swarm is fed back to the velocity update formula of the particle swarm.(2)Modify the framework of the artificial fish swarm algorithm,and add a repair mechanism to solve the multi-period portfolio model with cardinality constraints.(3)A new multi-objective adaptive chaotic particle swarm optimization algorithm is proposed to simulate the chaotic stability of particle swarms.Alternate processes,and adaptive adjustment of inertia weighting strategy.(4)Proposed a new step-by-step tolerance method,layering the objective function according to the subjective preferences of investors.(2)Consider the fundamental information of the asset in the portfolio model to construct a mean-variance-efficiency portfolio model.At present,most portfolio models only consider the historical rate of return on securities assets,and do not use other information of securities assets,such as the fundamental information of securities assets.In the real and complex securities market,investors often want to make the use of all kinds of information.Avoid yourself being in information that is not equal.To this end,this paper blurs some of the fundamental indicators of securities assets,and uses the fuzzy DEA model to obtain the efficiency of each stock,thus establishing a mean-variance-efficiency portfolio model.(3)Consider a portfolio model that is closer to reality.In this paper,we consider the factors of portfolio investment,transaction cost,risk,skewness and cardinality constraints to study the multi-period portfolio selection with cardinal constraints in fuzzy environment.Based on the possibility theory,a multi-period portfolio model with base-constrained income-risk and return-risk-skewness is established.(4)Using the dual hesitation fuzzy theory to study the portfolio problem in the case of insufficient information.The use of dual hesitant fuzzy sets to characterize securities assets has not yet been addressed.In this paper,we firstly study the variance function of the even hesitant fuzzy set,then measure the return and risk of the portfolio with the score function value and the variance function value of the dual hesitant fuzzy set,and establish the portfolio model under the dual hesitant fuzzy environment.Finally,a concrete numerical example is given.The particle swarm optimization algorithm is used to solve the model and prove the validity of the proposed model.