Optimizing permutation-based problems with a discrete vine-copula as a model for EDA

Abdelhakim Cheriet, Roberto Santana

Abstract: In this paper, we introduce a copula-based EDA that uses a Discrete Vine-Copula (DVC) model. This model is particularly suited to represent distributions in the permutation representation. To demonstrate the effectiveness of the proposed Discrete-Vine-Copula based EDAs (DVCEDA), we perform a set of experiments on instances of the known TSP problems. The results obtained are promising to extend the work on other class of problems.