Dynamic potential-based reward shaping.

Sam Devlin, Daniel Kudenko

Browse the full AAMAS paper archive.