二阶随机占优约束优化问题的遗传算法求解

Solution to constrained optimization problem of second-order stochastic dominance by genetic algorithm

随机占优是经济学和决策论中的基本概念,在投资组合优化中得到了广泛的应用.遗传算法无须求解目标函数和约束函数的次微分,也不用满足Slater约束规范,解决了约束的半无限性和非光滑性等问题.两个算例表明,遗传算法能很好地解决投资组合优化问题,并且效率得到了很大提高.

Stochastic dominance is fundamental concept in economics and decision-making theory, and is widely applied to portfolio optimization in recent years. Genetic algorithm has the advantages, which doesn＇t need to solve the subdifferential of Slater constrained rules, so it can solve the cons object trained function and constrained function or to satisfy semi-infinite and non-smooth problem. Two examples show that the genetic algorithm can well solve the portfolio optimization problem and the efficiency is greatly improved.