Material Detail

Active Reinforcement Learning

Active Reinforcement Learning

This video was recorded at 25th International Conference on Machine Learning (ICML), Helsinki 2008. When the transition probabilities and rewards of a Markov Decision Process (MDP) are known, the agent can obtain the optimal policy without any interaction with the environment. However, exact transition probabilities are difficult for experts to specify. One option left to an agent is a long and potentially costly exploration of the environment. In this paper, we propose another alternative: given initial (possibly inaccurate) specification of the MDP, the agent determines the sensitivity of the optimal policy to changes in transitions and rewards. It then focuses its exploration on the regions of space to which the optimal policy is most sensitive. We show that the proposed exploration strategy performs well on several control and planning problems.


  • User Rating
  • Comments
  • Learning Exercises
  • Bookmark Collections
  • Course ePortfolios
  • Accessibility Info

More about this material


Log in to participate in the discussions or sign up if you are not already a MERLOT member.