Learning First-Order Symbolic Representations for Planning from the Structure of the State Space.

Blai Bonet Hector Geffner

Published in: ECAI (2020)

Keyphrases

state space
symbolic representation
stochastic domains
reinforcement learning
learning algorithm
training data
heuristic search
macro actions
active learning
planning problems
feature selection
prior knowledge
first order logic
action selection
partially observable