Artificial Intelligence Course Projects

Overview

This repository contains a curated collection of artificial intelligence projects implemented through practical programming exercises. The projects cover foundational AI topics including game-based learning, neural-network image classification, heuristic search, A* search, state-space modeling, and reproducible experiment documentation.

Technical Focus

Artificial intelligence combines algorithmic problem solving, data-driven modeling, and structured decision-making. The projects in this repository demonstrate how AI problems can be represented, implemented, evaluated, and documented using Python-based workflows.

Core themes include:

• Game-state modeling and rule-based simulation
• Reinforcement-style learning through repeated gameplay
• Neural-network and convolutional-neural-network implementation
• Image classification with TensorFlow/Keras workflows
• Heuristic search and A* path finding
• State-space representation and transition modeling
• Reproducible experiment organization
• Clean technical documentation for public repositories

Skills & Technologies

Core skills: Python, Artificial Intelligence, Search Algorithms, Heuristic Search, A* Search, State-Space Modeling, Game AI, Reinforcement-style Learning, TensorFlow, Keras, CNN, Image Classification, Computer Vision, Jupyter Notebook, Experiment Documentation

Project Directory

No.	Project	Topic	Purpose	Main Concepts	Skills / Tags
01	Six-Pawn Reinforcement Learning	Game-based AI learning	Implement a six-pawn game learner that improves move selection through repeated play and stored learning values.	Game state representation, legal moves, decision policy, JSON-based learning data, simulation loop	Python, Game AI, Simulation, Reinforcement-style Learning, JSON
02	Deep Learning Image Classification	Neural networks and CNNs	Implement dense and convolutional neural-network workflows for image classification experiments.	MNIST/CIFAR-style classification, dense layers, Conv2D, Flatten, parameter counting, training-curve interpretation	Python, TensorFlow, Keras, CNN, Computer Vision, Jupyter
03	A* Puzzle Solving	Heuristic search and puzzle solving	Implement A* search for structured puzzle-solving problems using explicit state-space modeling.	State representation, transition generation, admissible heuristics, priority queue, optimal-path reconstruction	Python, A* Search, Heuristic Search, Algorithms, State-Space Modeling

Skill Coverage Map

Skill / Concept	01 Six-Pawn Learning	02 Image Classification	03 A* Puzzle Solving
Python programming	✓	✓	✓
Artificial intelligence problem modeling	✓	✓	✓
Game simulation	✓
Reinforcement-style learning	✓
JSON-based experiment data	✓
Jupyter Notebook workflow		✓
Neural networks		✓
Convolutional neural networks		✓
Image classification		✓
Computer vision workflow		✓
Heuristic search			✓
A* search			✓
State-space modeling	✓		✓
Priority queue / heap usage			✓
Result documentation	✓	✓	✓

Project Highlights

01. Six-Pawn Reinforcement Learning

This project implements a simple game-based AI learner for a six-pawn board game. The learner improves move selection through repeated gameplay and stored learning values.

Key capabilities include:

• Representing board states and legal actions
• Simulating repeated game sessions
• Updating move preferences based on outcomes
• Persisting learning data in JSON format
• Documenting experiment behavior and results

02. Deep Learning Image Classification

This project implements neural-network and convolutional-neural-network workflows for image classification. It includes model construction, notebook-based experimentation, architecture analysis, and summarized training observations.

Key capabilities include:

• Building dense neural-network models
• Building CNN models with convolutional layers
• Applying TensorFlow/Keras training workflows
• Comparing model structures and parameter counts
• Interpreting image-classification experiment outputs

03. A* Puzzle Solving

This project implements A* search for classic puzzle-solving tasks, including river-crossing and vessel-pouring style problems. It focuses on formal state representation, valid transition generation, heuristic design, and optimal-path reconstruction.

Key capabilities include:

• Modeling puzzle states and transitions
• Designing and applying heuristic functions
• Using priority queues for frontier management
• Reconstructing solution paths
• Handling unsolved or invalid search cases

Repository Structure

artificial-intelligence-course-projects/
├── README.md
├── .gitignore
├── 01-six-pawn-reinforcement-learning/
│   ├── README.md
│   ├── src/
│   ├── data/
│   ├── scripts/
│   └── results/
├── 02-deep-learning-image-classification/
│   ├── README.md
│   ├── notebooks/
│   ├── src/
│   ├── assets/
│   ├── scripts/
│   └── results/
└── 03-astar-puzzle-solving/
    ├── README.md
    ├── src/
    ├── scripts/
    ├── assets/
    └── results/

Some project folders may omit directories that are not required for that specific implementation.

Reproducibility Notes

Each project folder includes its own README.md with project-specific setup and execution instructions. Typical Python-based projects can be run with:

python3 src/<script_name>.py

Notebook-based experiments can be opened with:

jupyter notebook

Recommended environment setup:

python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt

If a project does not include a shared requirements.txt, install only the dependencies listed in that project's README.