A genetic algorithm for a portfolio optimization model
Gong Xue-wen · Journal of Northwest Normal University · 2012
A mean-absolute deviation portfolio optimization model with transaction cost is established,which also take account of several constraints,such as integer transaction,investment limit and total capital.A stochastic repair method of unfeasible solutions is proposed according to the specific characteristics of the solutions,and then a new genetic algorithm based the repair of unfeasible solutions is developed.The empirical analysis shows that the genetic algorithm has higher search efficiency and stability.By setting appropriate genetic generation,starting from arbitrary initial population,the algorithm can obtain an approximate optimal solution of the problem stably.