Skip to content

Latest commit

 

History

History
20 lines (11 loc) · 609 Bytes

README.md

File metadata and controls

20 lines (11 loc) · 609 Bytes
Raw

Source from :

POMCP 1.0

This is the release of POMCP used in the NIPS 2010 paper "Online Monte-Carlo Planning in Large POMDPs" by David Silver and Joel Veness

Added Code

We introduce human knowledge palyer into the game by introducing new modes for choosing of legal actions at each stage. We show this human knowledge can contribute to better policy design than that of the base line model

Future work

  • The current human knowledge works only on cells of distance one from the node we plan to introuc more complex examples
  • We plan to do more in depth experiments with less restrictions