A collection of miniprojects developed for a university course covering various topics in Artificial Intelligence.
- 01-pathfinding-search: Implements and visualizes heuristic search algorithms (such as Greedy Best-First Search with Manhattan, Euclidean, and Chebyshev heuristics) for grid-based pathfinding in a 2D maze.
- 02-connect-four-minimax: Features an interactive Connect Four game with an AI opponent powered by the Minimax algorithm with alpha-beta pruning and custom heuristic board evaluation.
- 03-genetic-algorithm: Implements a Genetic Algorithm (GA) from scratch in Python to find the global minimum of the 2D Himmelblau's function and evaluates the impact of its key hyperparameters.
- 05-neural-network-from-scratch: A pure NumPy implementation of a Multilayer Perceptron (MLP) trained and evaluated on the Fashion-MNIST dataset, featuring custom dense layers and activation functions.
- 06-reinforcement-learning-q-learning: An implementation of the off-policy Q-learning algorithm trained on the discretized Gymnasium MountainCar-v0 environment, examining learning rate, discount factor, and exploration decay parameters.
- 07-prolog-date-interval: A Prolog program that parses, validates, and calculates the absolute number of days between two dates within the 2024 leap year.