Estimation Of Distribution Algorithm Based On Copula And Its Application In Financial Risk Analysis

Copula theory provides a new method for obtaining joint distribution, and proposes to decompose the joint distribution into a continuous function and multiple marginal distributions, of which the marginal distribution reflects the univariate change. In addition, the estimation of distribution algorithm has the advantage of high efficiency in operation, but its operation is more complicated and the operand is relatively large in the case of estimation of probability model. Therefore, first of all, this paper combines Copula theory and estimation of distribution algorithm; makes use of the advantages of Copula theory to simplify the estimation of distribution algorithm and establish the process of probability distribution model; applies the Copula estimation of distribution algorithm to the field of financial risk analysis, and introduces the duplication thought of flora algorithm for improvement. Then, it compares the calculation results of Copula-Va R model with the risk measurement results of Copula estimation of distribution algorithm, in order to prove the effectiveness of the algorithm application. In terms of the selection of algorithm objective optimization function, this paper introduces the objective function of risk adjusted return on capital, which better complies with the actual meaning of risk.Finally, this paper conducts empirical description that the Copula function can get the effective information between different financial assets, especially the characteristics that financial asset does not follow the normal distribution and has the fat-tail distribution characteristics. In the face of the increasing risk of financial market, investors will put forward more and more strict requirements for risk measurement analysis, because they want to maximize the benefits with less risk, so as to better conduct asset allocation. Therefore, it is of practical significance to apply the Copula estimation of distribution algorithm to risk analysis, which provides a method for risk measurement.