Ant Colony Paradigms

Michael R. Gargano, Pace University, Computer Science and Information Systems Department

Document Type Article

This paper was officially published as a Technical Report paper for the Computer Science and Information Systems department of Pace University in June, 2002.

Abstract

Ant colony (or swarm intelligence)optimization has become and important and powerful optimization heuristic in computer science. This thesis will describe some models which can be used to solve various interesting problems. Through several examples and experiments this thesis will validate the benefits and potentials of this technique.

This thesis begins by studying a construction model and then a random walk problem with potential applications to data compression. The second half studies three classical optimization problems: graph partitioning (GP)(for data categorization), the traveling salesperson problem (TSP), and the minimum quadratic assignment problem (QAP)