Reward Propagation
Using Graph Convolutional Networks
Martin Klissarov Doina Precup
Mila, McGill University Mila, McGill University and DeepMind
martin.klissarov@mail.mcgill.ca dprecup@cs.mcgill.ca
Abstract
Potential-based reward shaping provides an approach for designing good reward
functions, with the purpose of speeding up learning. However, automatically find-
ing potential functions for com ...


雷达卡



京公网安备 11010802022788号







