Local Policy Optimization for Trajectory-Centric Reinforcement Learning.
Patrik KolaricDevesh K. JhaArvind U. RaghunathanFrank L. LewisMouhacine BenosmanDiego RomeresDaniel NikovskiPublished in: ICRA (2020)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- optimization problems
- global optimization
- reinforcement learning algorithms
- markov decision process
- action selection
- machine learning
- partially observable
- markov decision processes
- reinforcement learning problems
- optimization process
- partially observable environments
- genetic algorithm
- model free
- state space
- real robot
- control policies
- approximate dynamic programming
- policy evaluation
- partially observable domains
- state and action spaces
- transition model
- actor critic
- neural network
- markov decision problems
- reward function
- temporal difference
- constrained optimization
- function approximation
- optimization method
- transfer learning
- multi agent