University of Cambridge > Talks.cam > Engineering Safe AI > Safe Exploration in Reinforcement Learning

Safe Exploration in Reinforcement Learning

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact AdriĆ  Garriga Alonso.

In reinforcement learning, the learning agent always faces a tradeoff between exploration and exploitation. Often, exploration is implemented as taking random actions, but in dangerous tasks, this can lead to highly negative rewards – damage to the agent or other parts of the environment. This week we discuss methods to minimize the risks in exploration and learn good policies both safely and efficiently.

Slides: https://valuealignment.ml/talks/2018-01-31-safe-exploration.pdf

This talk is part of the Engineering Safe AI series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2018 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity