Elastic Neural Nets for the Traveling Salesman Problem

Article type: Research Article

Authors: Antoniou, I.^{a; b} | Gorbunov, S.V.^c | Ivanov, V.V.^{a; d; 1} | Kisel, I.V.^{c; e} | Konotopskaya, E.V.^c

Affiliations: [a] International Solvay Institutes for Physics and Chemistry, CP-231, ULB, Bd. du Triomphe, 1050, Brussels, Belgium | [b] Theoretische Natuurkunde, Free University of Brussels, Brussels, Belgium | [c] Joint Institute for Nuclear Research, 141980, Dubna, Russia | [d] Laboratory of Information Technologies, Joint Institute for Nuclear Research, 141980, Dubna, Russia | [e] Max-Planck-Institut für Physik, Werner Heisenberg Institut, Föhringer Ring 6, D-80805 München, Germany

Correspondence: [1] The address for manuscript correspondence: V.V. Ivanov, Laboratory of Information Technologies, Joint Institute for Nuclear Research, 141980, Dubna, Russia. Fax: +7 09621 65145. E-mail: [email protected].

Abstract: In order to speed up the standard Elastic Neural Net (ENN) [1] and make it robust against, local minima problem, we present three variants of the ENN, namely, the Modified ENN, the Double ENN and the Smooth ENN. The efficiency of the new algorithms is demonstrated and compared with some well-known classical algorithms on collections of cities with known optimal tours.

Keywords: Traveling salesman problem, Elastic neural net

DOI: 10.3233/JCM-2002-21-215

Journal: Journal of Computational Methods in Sciences and Engineering, vol. 2, no. 1-2, pp. 111-115, 2002