LP-GNN learns vectorized planning domain models via GNNs from partial traces and outperforms the ARMS learner on solving problems across five classical domains.
Domain model acquisition in domains with action costs
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Representation Learning for Classical Planning from Partially Observed Traces
LP-GNN learns vectorized planning domain models via GNNs from partial traces and outperforms the ARMS learner on solving problems across five classical domains.