ModGNN: Expert Policy Approximation in Multi-Agent Systems with a Modular Graph Neural Network Architecture.
Ryan KortvelesyAmanda ProrokPublished in: CoRR (2021)
Keyphrases
- multi agent systems
- directed graph
- multi agent
- inverse reinforcement learning
- autonomous agents
- game theory
- graph theory
- expert knowledge
- approximation algorithms
- graph representation
- graph model
- cooperative
- coalition formation
- weighted graph
- graph structure
- bipartite graph
- closed form
- optimal policy
- connected components
- structured data
- agent architecture
- graph mining
- undirected graph
- graph theoretic
- domain knowledge
- domain experts
- directed acyclic graph
- graph databases
- random walk
- markov chain
- adjacency matrix
- min sum