On computation methods of the minimax regret solution for linear programming problems with uncertain objective function coefficients

M. Inuiguchi, H. Higashitani, T. Tanino
IEEE SMC'99 Conference Proceedings. 1999 IEEE International Conference on Systems, Man, and Cybernetics (Cat. No.99CH37028)  
In this paper, we investigate the computation methods for minimax regret solutions to linear programming problems with uncertain objective function coefficients. The previously proposed solution algorithms, two phase approach, bilevel programming approach and branch and bound approach are reviewed. An outer approximation approach is newly proposed. By numerical experiments, the efficiency of solution algorithms is compared. It is shown that a computation method based on the branch and bound
more » ... ranch and bound approach is fastest in interval coefficient case and that the outer approximation approach is effective as the problem size increases and can be fastest in polytope case among the three approaches.
doi:10.1109/icsmc.1999.823361 fatcat:oxykota43jc2nhol5o5qgccneu