Intellectics Group: Technical Report 97-04
MAX-MIN Ant System for Quadratic Assignemnt Problems
In this paper we present the application of MAX--MIN Ant System to the Quadratic
Assignment Problem. MAX--MIN Ant System is an improvement over Ant System,
an earlier proposed algorithm of Ant Colony Optimization. The computational results
show that very high solutions quality can be obtained using our approach.
As MAX--MIN Ant System is a hybrid algorithm combining
solution construction with local
search, the kind of local search procedure has significant influence on
its final performance. In particular, we show that the best choice of a
local search procedure depends strongly on the instance type.