Planning with Learned Object Importance in Large Problem Instances using Graph Neural Networks.
Tom SilverRohan ChitnisAidan CurtisJoshua B. TenenbaumTomás Lozano-PérezLeslie Pack KaelblingPublished in: AAAI (2021)
Keyphrases
- neural network
- pattern recognition
- structured data
- graph representation
- object model
- graph theory
- planning problems
- complex objects
- planning process
- random walk
- fuzzy logic
- artificial neural networks
- subgraph isomorphism
- neural network model
- back propagation
- spatial relations
- weighted graph
- genetic algorithm
- category specific
- causal graph
- graph theoretic
- object models
- graph partitioning
- spanning tree
- graph mining
- object segmentation
- neural nets
- spatial relationships
- heuristic search
- reinforcement learning