Neural Nets WIRN Vietri-01 pp 87-95
Solving Min Vertex Cover with Iterated Hopfield Networks
- Cite this paper as:
- Bertoni A., Campadelli P., Grossi G. (2002) Solving Min Vertex Cover with Iterated Hopfield Networks. In: Tagliaferri R., Marinaro M. (eds) Neural Nets WIRN Vietri-01. Perspectives in Neural Computing. Springer, London
A neural approximation algorithm for the Min Vertex Cover problem is designed and analyzed. This algorithm, having in input a graphs G = (V, E), constructs a sequence of Hopfield networks such that the attractor of the last one represents a minimal vertex cover of G. We prove a theoretical upper bound to the sequence length and experimentally compare on random graphs the performances (quality of solutions, computation time) of the algorithm with those of other known heuristics. The experiments show that the quality of the solutions found by the neural algorithm is quite satisfactory.
KeywordsMin Vertex Cover Hopfield networks approximation algorithms
Unable to display preview. Download preview PDF.