CAP6635 Artificial Intelligence

A summary of the AI techniques explored in Dr. Zhu's AI class

Structure of Package

• cap6635.agents define the different search and decision-making algorithms.
  • See sub-directories for more details on the agents
• cap6635.environment creates and populates the world with various obstacles or other states.
• cap6635.utilities hosts various helper functions for searching, environment manipulation, animation and more.

Development setup

pip install -r requirements.txt cap6635

AI Examples

Vacuums

Run the vacuums.py example with the optional paramters. The output gets saved as a vacuum?.gif animation.

# Type of agent defaults to random type (if not provided)
# 1 --> Simple Reflex Vacuum
# 2 --> Model-based Vacuum
# 3 --> Goal-based Vacuum
# World Height & Width defaults to random int (if not provided)

python 1_vacuums.py [type_of_agent] [height_of_world] [width_of_world]

n-Queens

# Hill Climbing
python 2_hill_climbing.py [number_of_queens]

# Simulated Annealing
python 3_simulated_annealing.py [number_of_queens]

# Genetic Algorithm
python 4_genetic.py [number_of_queens]

Tic Tac Toe

# Minimax + Alpha-Beta Pruning
# Algorithm {1 --> Minimax, 0 --> Alpha-Beta Pruning}
# First Player {1 --> Human, 2 --> AI}
python 5_minimax.py [Algorithm] [First Player]

# e.g. Minimax (Alpha-Beta Pruning) - Player is X
python 5_minimax.py 0 1