I have a question about the MIP solution process, something I observed during a simulation. In the following image, you see that the incumbent objective bound (which I understand as the upper bound of the objective) is not updated even when the current node has a better objective value in this minimization problem. This I don't understand. To me, if there is already a better solution that is feasible, then that should be the upper bound to the objective value. Can anyone explain to me this behavior?
Please sign in to leave a comment.