Improving the Exploration of Deep Reinforcement Learning in Continuous Domains using Planning for Policy Search.
Jakob J. HollensteinErwan RenaudoMatteo SaverianoJustus H. PiaterPublished in: CoRR (2020)
Keyphrases
- continuous domains
- policy search
- reinforcement learning
- action selection
- reinforcement learning algorithms
- continuous state
- function approximation
- partially observable markov decision processes
- state space
- partially observable
- learning algorithm
- machine learning
- temporal difference
- dynamic programming
- model free
- policy gradient
- reinforcement learning methods
- multi agent
- reward function
- markov decision process
- markov decision processes
- optimal policy
- evolution strategy
- action space
- optimal control
- policy iteration
- function approximators
- transfer learning
- markov decision problems
- heuristic search
- genetic algorithm