Title:
SwarmBot
Author(s):
Christopher Blaha
Department:
Computer Science
Faculty Mentor(s):
Marietta Cameron and Kenneth Bogert
Abstract / Summary:
Swarm intelligence utilizes multiple agents that behave like a swarm of insects or other animals to optimize the skill of an AI program. Ants follow a pheromone to the optimal path, fireflies group towards the brightest bug, and cuckoo birds lay their eggs in other bird’s nests to pass off as the other bird's babies. Those that survive pass on a genetic algorithm that dictates behavior towards the optimal path. Multiple AI agents using these behaviors simultaneously have an advantage over a single agent working alone. They can cover more ground when searching for the winning strategy. They also have a considerably faster performance compared to a single agent. This project utilizes the ant colony, firefly, and cuckoo bird swarm methods to compare performance in the game Othello.
Publication Date:
Jan-9-2024
Documents: