There are many ways to approach optimization. In metaheuristics, an algorithm generally operates on candidate solutions without making any assumptions about the problem space. Algorithms are designed to search large, many-dimensional spaces, and to avoid getting stuck at local extrema.

Commonly discussed forms of metahueristics include the genetic algorithm, simulated annealing, Tabu search, genetic programming, particle swarm optimization, and ant colony optimization.

For more information about the field of metaheuristics, see these resources:
Metaheuristic (Wikipedia): Provides a good summary of the many algorithms developed so far.
Essentials of Metaheuristics: This book is an excellent resource for novices to the field.

Last edited Sep 23, 2011 at 5:28 PM by jamestunnell, version 4


No comments yet.