Randomized Local Extrema for Heuristic Selection in TSP

Abstract

It follows from the search randomizations in space-time among candidate heuristics that the optimality of an arbitrary heuristic is unsolvable. There are a countable infinite number of theories that may be decomposed into stronger local proofs. Local inductive randomization depends on domain symmetry for tractability. TSP problems exhibit tentative domain symmetry and potential space-time randomness in domain solution evolution. Heuristics in the domain of the TSP can be found and selected with a suitable representation, randomization, and symmetric induction with a significantly reduced time. Better representation of the TSP problem facilitates a better solution