Starting from October 2009, I am a Lecturer (Assistant Professor) at the Department of Information Systems Engineering at Ben Gurion University, Israel.

Previously, I was a researcher at Microsoft Research from 2007 to 2009.

I graduated in 2007 from Ben Gurion University, Israel, under the supervision of Prof. Ronen Brafman and Prof. Solomon E. Shimony. My dissertation has focused on learning and solving Partially Observable Markov Decision Processes (POMDPs), including both efforts in Reinforcement Learning under partial observability, and solving POMDPs using Point-Based methods. My M.Sc. dissertation, also under the supervision of Prof. Brafman and the guidance of David Heckerman, covered the application of a Markov Decision Process to a Recommender System.

• 
  Ph.D. 2003-2007, Ben-Gurion University, Dept. of Computer Science.
• 
  Thesis title: Learning and Solving POMDPs
• 
  Advisor: Prof. Eyal S. Shimony and Prof. Ronen I. Brafman.
• 
  M.Sc. 1999-2002, Ben-Gurion University, Dept. of Computer Science
• 
  Thesis title: An MDP-Based Recommender System
• Advisor: Prof. Ronen I. Brafman a - David Heckerman.
• B.Sc. 1996-1999, Ben-Gurion University, Dept. of Computer Science

• My main research field is control under uncertainty. I am mainly interested in POMDPs, but also in any other technique that can be used to provide optimal control when actions have stochastic effects and the world features are not fully available. I am interested both in a model-based reinforcement learning, where a model of the environment is learned through trial and error, or through previous recorded interactions with the environment, and in obtaining approximate solutions to such models. My current goal is to find good applications of POMDP techniques to real world problems. Such success stories can help the POMDP community to make a considerable impact on the research community.

• I am also active in the recommender systems research. Recommender systems attempt to suggest items to users, usually customers of an e-commerce web site, trying to enrich the user experience but also to provide value to the site owner. I am interested in new methods to create predictive models of user behavior, in the mining of relevant data for recommendations, in advanced methods for measuring the quality of recommendations, and in any other topic concerning recommender systems.

• I am interested in applications of Machine Learning techniques to real world problems. Specifically, I am mostly interested in problems that arise in the software industry, rather than in the physical world. I believe that there is much potential to apply ML techniques for understanding, managing, and controlling software processes.