Carleton University - School of Computer Science Honours Project
Winter 2017
The Ant Colony Optimization Heuristic Applied to the Protein Folding Problem
ABSTRACT
The study of protein folding has led to and will continue to provide many breakthroughs
in the fields of health, medicine, and agriculture. A process of determining the lowest energy
conformation of a protein sequence is often aided by computational optimization strategies
such as the Ant Colony Optimization metaheuristic. Changes were made to how the ACO
metaheuristic behaves in an attempt to improve solution quality and time efficiency. The
results strongly indicate no improvements were made as acceptable energy states and run
times were not achieved. In conclusion, normalization of pheromone and heuristic values is of
the highest priority when implementing an ACO metaheuristic. Scaling factors associated with
pheromone and heuristic values can be fine tuned to provide significant increases in
optimization efficiency.