Home page of Shie Mannor

My research interests lie in the union of Machine learning, Systems and Control, and Game Theory.

• Adaptive control and learning algorithms for dynamical systems, specifically Reinforcement Learning  (aka Neuro-Dynamic Programming).
• Machine Learning (theory, algorithms, and applications). I am most interested in high-dimensional problems with uncertainty in the data.
• Markov decision processes - theory and applications; particularly large scale Markov processes that require methods like approximate dynamic programming.
• Stochastic, dynamic, network, and differential games; applications in communication networks.
• Multi-agent systems, and especially learning in such systems (e.g., Universal Consistency and Regret Minimization).
• Optimization of large scale problems and especially combinatorial optimization using heuristic methods (e.g., the Cross Entropy method) and stochastic optimization.

Teaching

• Fall 2004, Fall 2005, Fall 2006 and Fall 2007: Introduction to Software Engineering (ECSE-321); access restricted to McGill community by WebCT.
• Fall 2005, Fall 2006 and Fall 2007: Multi-agent systems (ECSE-508); access restricted to McGill community by WebCT.

Message:

1. Graduate students are always welcome! (please read this, and please don't email me directly - you have to apply through the McGill admission office - I cannot answer all the emails I receive from prospective students and I will ignore your email).
    

How to pronounce my name: Shie is pronounced like "shy", and Mannor rhymes with "Amor".

Contact

McGill University
Department of Electrical & Computer Engineering
McConnell Engineering Building, Rm. 525
3480 University Street,
Montreal, Quebec,
Canada H3A-2A7
Tel. 514-398-1467, Fax. 514-398-4470

E-mail: My first name DOT my last name AT mcgill DOT ca