An improved optimization algorithm was designed for finding these solutions of discontinuous portfolio optimization models in stock market quickly and efficiently. By introducing crossover operations, an innovative particle swarm optimization algorithm (CPSO) based on optimal and sub-optimal locations was proposed. Then in performance test the algorithm performed better than some existed improved particle swarm optimization algorithms, and overcome the flaw of premature. Finally, CPSO was applied in stock market, and in simulation experiment optimization values of two portfolio models under different expected return rates were obtained.

particle swarm optimization algorithm, crossover operation, portfolio, expected return rate, stock market.