The traveling salesman problem (TSP) is to find the
shortest hamiltonian cycle in a graph. This problem is
NP-hard and thus interesting. There are a number of
algorithms used to find optimal tours, but none are
feasible for large instances since they all grow exponentially.
We can get down to polynomial growth if we settle
for near optimal tours. We gain speed, speed and speed
at the cost of tour quality. So the interesting properties
of heuristics for the TSP is mainly speed and closeness
to optimal solutions.
There are mainly two ways of finding the optimal
length of a TSP instance. The first is to solve it optimally
and thus finding the length. The other is to
calculate the Held-Karp lower bound, which produces
a lower bound to the optimal solution (see section 7).
This lower bound is the de facto standard when judging
the performance of an approximation algorithm for
the TSP.
The heuristics discussed here will mainly concern
the Symmetric TSP, however some may be modified to
handle the Asymmetric TSP. When I speak of TSP I
will refer to the Symmetric TSP.