Prioritized Soft Q-Decomposition for Lexicographic Reinforcement Learning.
Finn RietzErik SchaffernichtStefan HeinrichJohannes A. StorkPublished in: ICLR (2024)
Keyphrases
- reinforcement learning
- function approximation
- decomposition method
- markov decision processes
- learning algorithm
- image decomposition
- temporal difference
- state space
- optimal policy
- learning process
- optimal control
- reinforcement learning algorithms
- multi agent
- function approximators
- robotic control
- dynamic programming
- stochastic approximation
- reinforcement learning methods
- temporal difference learning
- cardinality constraints
- machine learning
- search algorithm
- action selection
- model free
- search space
- learning problems
- combinatorial optimization