Solving portfolio selection problem using Dantzig-Wolfe algorithm

Portfolio selection process is one of the problems that have been attracted many researchers. Various criteria that have been applied in this case have changed over time and this situation makes necessary the using of appropriate tools to support investment decisions. The purpose of this research is modeling and solving of portfolio selection problem. On the other hand, in some cases of a portfolio optimization, due to largeness of problem size, the problem would be impossible to solve in a reasonable time. In such situation, applying the methods that reduce the scale of problem can be useful. In current paper a Dantzig-Wolfe algorithm is used to solve the problem in which, after decomposing the basic problem into several sub problems and solving them, individually, the obtained results are aggregated. The results of applying this method showed its efficiency in solving the large-scale problems show.