When is the Assignment Bound Tight for the Asymmetric Traveling-Salesman Problem?
SIAM Journal on Computing
The probabilistic relationship between the assignment and asymmetric traveling salesman problems
SODA '01 Proceedings of the twelfth annual ACM-SIAM symposium on Discrete algorithms
Hi-index | 0.00 |
We study the relationship between the value of optimal solutions to the random asymmetric b-partite traveling salesman problem and its assignment relaxation. In particular we prove that given a bn×bn weight matrix W = (wij) such that each finite entry has probability pn of being zero, the optimal values bATSP(W) and AP(W) are equal (almost surely), whenever npn tends to infinity with n. On the other hand, if npn tends to some constant c then P[bATSP(W) ≠ AP(W)] Ɛ 0, and for npn → 0, P[bATSP(W) ≠ AP(W)] → 1 (a.s.). This generalizes results of Frieze, Karp and Reed (1995) for the ordinary asymmetric TSP.