Theoretical and Experimental Analysis of the Jurema Branch-and-Bound Method
Felipe Martins Müller, Guilherme Dhein, Olinto César Bassi de Araújo · 2011
Jurema was presented as a new branch-and-bound method to find optimal solutions or to prove the optimality of a solution. In this work, we present a discussion about some aspects of the Jurema method. We show that the objective function value of the initial solution used by Jurema to guide the search process is less important than its structure. We also demonstrate that upper bounds closer to the optimal cost improve the search performance, but reduce or eliminate any Jurema's advantage over other branch-and-bound methods. Finally, an alternative implementation for the Jurema method, different from the originally proposed, is presented.