Section: New Results

Neutrality Analysis is Graph coloring problem

Participants: Aymeric Blot, Clarisse Dhaenens, Laetitia Jourdan, Marie-Eleonore Marmion

Solving neutral problems is challenging as many optimization methods have difficulty to obtain good solutions. Hence, studying the neutrality in order to provide insights on the structure of the problem to be soved may be an answer. This has been done for the graph coloring problem (GCP) for which the neutrality of some hard instances has been quantified. This neutrality property has to be detected as it impacts the search process. Indeed, local optima may belong to plateaus that represent a barrier for local search methods. Then, we also aim at pointing out the interest of exploiting neutrality during the search. Therefore, a generic local search dedicated to neutral problems, NILS, is performed on several hard instances [78] .