PolicyBlocks: An Algorithm for Creating Useful Macro-Actions in Reinforcement Learning

Marc Pickett, Andrew G. Barto · 2002

We present PolicyBlocks, an algorithm by which a reinforcement learning agent can extract useful macro-actions from a set of related tasks. The agent creates macroactions by finding commonalities in solutions to previous tasks. Using these macro-actions, learning to do future related tasks is accelerated.

Read the paper · More papers on PaperTik