Research on investment decision of portfolios at given risks

In this paper we apply accelerating genetic algorithm(AGA)to solve the model of investment decision of portfolios at the given risks in the light of characteristic of nonsatisfaction of investment. Studies show that AGA is strongly adapt to variation of searching space, which can avoid some shortcomings such as complicated calculation, premature convergence, and has clear criterion of setting technique of controlling parameters, has no requirement of semi-canonicity to covariance matrix of n ranks, can expand objective function extensively, and also can enhance forecasting precision.