GOING FURTHER INTO MODELING PROBABILISTIC COMBINATORIAL OPTIMIZATION USING COMMUNICATION CHANNEL

This work builds upon a previous novel approach developed by N. Ngambou. Djakam and Murat. M. Tanik to tackle combinatorial optimization problems in which uncertainty is present as a significant concept. It is an approach akin to Herbert Simon's notion of “satisficing” in the context of bounded rationality. This paper demonstrates that a probabilistic maximum independent set(PIS) as a bipartite graph can be represented as a communication channel and explore a technique to find satisficing solutions of probabilistic combinatorial problems. Our approach differs from the a priori optimization method proposed by D. J. Bertsimas, P. Jaillet, and A. R. Odoni in the sense that after determining the solution of the whole instance (a priori solution), instead of applying a modification strategy, we can derive a solution of any sub-instance from the communication channel model.