How do Chess Engines Work - Looking at Stockfish and AlphaZero

Overview

Explore chess AI evolution:

from Deep Blue's algorithms to Stockfish's Minimax, and AlphaZero's revolutionary self-learning approach using Monte Carlo Tree Search and neural networks.

Syllabus

• Introduction to Chess AI
Historical overview of chess engines
Importance and impact of AI in chess
• Deep Blue: The Foundation
Overview of Deep Blue's architecture
Rule-based systems and brute force search
Case study: Garry Kasparov vs. Deep Blue
• Stockfish and the Minimax Algorithm
Fundamentals of the minimax algorithm
Role of alpha-beta pruning
Heuristic evaluation functions
Strengths and limitations of Stockfish
Case study: Deep Dive into a Stockfish game
• Introduction to Machine Learning in Chess
Basics of machine learning and neural networks
Transition from rule-based to learning-based systems
• AlphaZero: A New Era
Overview of AlphaZero's architecture
Monte Carlo Tree Search explained
Reinforcement learning and self-play methods
Neural network training processes
Innovations and contributions to AI
• Comparing Stockfish and AlphaZero
The difference in approaches and architectures
Strengths, weaknesses, and practical applications
Performance analysis and comparative results
• Future Trends in Chess AI
Current research directions
Potential advancements and challenges
Ethical considerations and impact on human chess
• Conclusion
Summary of key insights
Future of AI in strategic games
• Course Wrap-up and Final Q&A Session

Subjects

Conference Talks