Reducing Planning Complexity of General Reinforcement Learning with Non-Markovian Abstractions.
Sultan Javed MajeedMarcus HutterPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- sequential decision problems
- special case
- action selection
- partially observable
- function approximation
- state space
- planning problems
- markov decision processes
- closely related
- heuristic search
- machine learning
- dynamic programming
- lower bound
- decision theoretic
- computational complexity
- multi agent
- partially observable markov decision processes
- markov decision process
- decision processes
- partially observable domains
- learning algorithm