eduzhai > Applied Sciences > Engineering >

Online inverse reinforcement learning with limited data

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 8 pages

Abstract: This paper addresses the problem of online inverse reinforcement learning forsystems with limited data and uncertain dynamics. In the developed approach,the state and control trajectories are recorded online by observing an agentperform a task, and reward function estimation is performed in real-time usinga novel inverse reinforcement learning approach. Parameter estimation isperformed concurrently to help compensate for uncertainties in the agent sdynamics. Data insufficiency is resolved by developing a data-driven update lawto estimate the optimal feedback controller. The estimated controller can thenbe queried to artificially create additional data to drive reward functionestimation.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...