Single-agent Planning# Classical Planning, Planning in Fully Observable Stochastic Environments, Planning in Partially Observable Stochastic Environments. Single-agent Reinforcement Learning (rl)# Tabular Methods Vs. Approximate Methods, Policy Gradient Vs. Value-based Methods, Model-based Vs. Model Free Rl. Multi-agent Planning# Planning in Adversarial, Cooperative, and Collaborative Multi-agent Settings. Communication and Resource Allocation In Multi-agent Systems. Multi-agent Rl# Learning in The Presence Of Others. Emergent Coordination and Cooperation. Efficient Communication In Noisy Partially Known Environments. Automated Design of Ai Environments to Promote Collaboration. Learning Outcomes# At The End of The Course The Students Will Know#knowledge of Various Ai Frameworks For Modeling Single-agent and Multi-agent Settings, Understanding The Theoretical Guarantees and Limitations of Different Ai Algorithms For Single and Multi-agent Planning and Rl, Acquiring Practical Experience Using Ai Tools and Implementing Them in Various Ai Domains, Experience in Analyzing The Performance of Different Ai Approaches, Constructing New Algorithms For Collaborative Ai.

Faculty: Computer Science
|Undergraduate Studies |Graduate Studies

Pre-required courses

236501 - Introduction to Artificial Intelligence or 236609 - Advanced Topics in Computer Science 9

Semestrial Information