Ant-FOIL: Integrating Ant Colony System and FOIL
Yan Chenyang · 2015
A novel and simple combination of inductive logic programming with swarm intelligence is presented. The Ant-FOIL tightly integrates the well-known inductive logic programming rule-learner FOIL with Ant Colony System meta-heuristic. The hypotheses construction is guided by the ACS stochastic local decision policy depends on pheromone and heuristic information. Experiments in applying Ant-FOIL to well-known benchmarks shows that Ant-FOIL performs better than either its baseline algorithm (FOIL), and is at the same time competitive with more sophisticated approaches (kFOIL, nFOIL and Aleph).