TRACE learns action embeddings from a transition model and transfers policies across reinforcement learning tasks with different state and action spaces, improving sample efficiency in experiments.
Jordan, and Philip S
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learning Action-Transferable Policy with Action Embedding
TRACE learns action embeddings from a transition model and transfers policies across reinforcement learning tasks with different state and action spaces, improving sample efficiency in experiments.