Lihong Li

[Education] [Interests] [Publications] [Software] [Professional Activities] [Working Experience] [Teaching Experience] [Collaborators]

• 01/2005 - 05/2009, PhD, Computer Science, Rutgers University, New Brunswick, NJ, USA
  • Advisor: Michael L. Littman
  • Thesis: A Unifying Framework for Computational Reinforcement Learning Theory
• 09/2002 - 07/2004, M.Sc., Computing Science, University of Alberta, Edmonton, AB, Canada
  • Advisors: Vadim Bulitko and Russell Greiner
  • Thesis: Focus of Attention in Reinforcement Learning
• 09/1998 - 07/2002, B.Eng., Computer Science and Technology, Tsinghua University, Beijing, P.R. China

• 06/2009 - present, postdoctoral scientist at Yahoo! Research
• 05/2008 - 08/2008, research intern at AT&T Shannon Labs (mentored by Jason D. Williams and Suhrid Balakrishnan)
• 05/2007 - 08/2007, research intern at Yahoo! Research NYC (mentored by John Langford)
• 05/2006 - 08/2006, engineer intern at Google NYC (mentored by Lee Laroco)
• 01/2005 - 06/2009, graduate research assistant at the Rutgers University
• 09/2002 - 07/2004, teaching/research assistant at the University of Alberta

• Vowpal Wabbit is a C++ implementation of two fast online-learning algorithms (with John Langford and Alex Strehl)
• MyAgent is a C++ implementation of reinforcement-learning agents that allows fast creation of many RL agents. It incorporates a number of techniques for function approximation, exploration, and value-function learning. (postponed)

• Reviewer for journals:
  • Journal of Machine Learning Research (2005 - 2010)
  • Journal of Artificial Intelligence Research (2008 - 2009)
  • Data Mining and Knowledge Discovery (2008)
  • Machine Learning (2009)
  • Neural Computation (2009)
  • Neurocomputing (2009)
  • IEEE Transactions on Wireless Communications (2009)
  • Journal of Computer Science and Technology (2009)
  • Artificial Intelligence Communications (2009)
  • IEEE Transactions on Neural Networks (2010)
• Reviewer or program committee for conferences:
  • AAAI: 2006, 2008, 2010
  • IJCAI: 2007
  • NIPS: 2008, 2009
  • ICML: 2009, 2010
  • ECML: 2009
• Program committeee for workshops:
  • IJCAI-05 workshop on planning and learning
  • ICML-09 workshop on abstraction in RL
• Funding Chair for the 2009 Reinforcement Learning Competition.

• In Spring 2009, I gave a guest lecture on least-squares policy iteration in Michael Littman's graduate class, "Learning and Sequential Decision Making"
• In Fall 2007, I helped organize a light seminar on Planning in Learned Environments
• In Summer 2005, I organized a seminar on Abstraction and Hierarchies for Learning and Planning
• From Fall 2002 to Winter 2004, I was a teaching assistant for discrete mathematics courses (Formal Systems and Logic in Computer Science), and won a University of Alberta UTS Teaching Award in 2004