LLQL: Logistic Likelihood Q-Learning for Reinforcement Learning.
Outongyi LvBingxin ZhouYu Guang WangPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- function approximation
- reinforcement learning algorithms
- model free
- state space
- optimal policy
- temporal difference learning
- state action space
- learning algorithm
- eligibility traces
- reinforcement learning methods
- action selection
- temporal difference
- maximum likelihood
- multi agent
- machine learning
- logistic regression
- continuous state and action spaces
- stochastic approximation
- real robot
- state action
- cooperative
- optimal control
- partially observable
- dynamic programming
- transfer learning
- markov decision processes
- learning process
- control problems
- rl algorithms
- learning problems
- continuous state
- multi agent reinforcement learning
- state abstraction
- policy search
- hierarchical reinforcement learning
- relational reinforcement learning
- reward function
- reward shaping
- learning capabilities