AI Algorithms Application: Optimization, Games, Logic & Classification

This repository contains the implementation and evaluation of various Artificial Intelligence (AI) algorithms to solve four distinct classes of real-world problems. The project emphasizes Object-Oriented Programming (OOP) principles and modular code structure using Python.

🚀 Key Features & Algorithms

1. Function Optimization (Simulated Annealing)

• Problem: Finding the global maximum of a complex mathematical function on a 3D surface within a defined range.
• Algorithm: Simulated Annealing.
• Key Implementation: Applied probabilistic acceptance of worse states to avoid local optima and successfully converge toward the global maximum.

2. Strategic Game (Tic-Tac-Toe 9x9)

• Problem: Developing an intelligent agent to play a 9x9 variant of Tic-Tac-Toe (Caro).
• Algorithm: Alpha-Beta Pruning integrated with H-minimax strategy.
• Key Implementation: Designed a custom Heuristic evaluation function to assess board states when the search tree depth is limited, ensuring strategic and efficient decision-making.

3. Logic Constraint (Propositional Logic & SAT Solver)

• Problem: Resolving map coloring and logic puzzles based on a set of given constraints.
• Algorithm: Propositional Logic transformation to Conjunctive Normal Form (CNF).
• Key Implementation: Utilized the PySAT library and SAT Solvers to model logical constraints and identify satisfiable solutions for complex combinatorial problems.

4. Machine Learning Classification (Naïve Bayes)

• Problem: Predicting student academic performance based on historical datasets.
• Algorithm: Naïve Bayes Classifier.
• Key Implementation: Performed Data Discretization to handle continuous variables and utilized conditional probability to categorize academic outcomes with high accuracy.

🛠 Tech Stack

• Language: Python
• Core Libraries: * SymPy: For symbolic mathematics and function processing.
• PySAT: For solving SAT-based logic problems.
• Matplotlib / NumPy: For data visualization and numerical computations.
• Pygame: For interactive game interface and visualization.

📈 Results & Conclusions

• Successfully implemented all algorithms with experimental results matching theoretical expectations.
• Developed a modular architecture that allows for easy maintenance and scalability of AI modules.
• Conducted detailed performance analysis regarding computational complexity and classification accuracy.