Virtual vs. real: Trading off simulations and physical experiments in reinforcement learning with Bayesian optimization.
Alonso MarcoFelix BerkenkampPhilipp HennigAngela P. SchoelligAndreas KrauseStefan SchaalSebastian TrimpePublished in: ICRA (2017)
Keyphrases
- reinforcement learning
- virtual space
- virtual environment
- virtual reality
- optimization problems
- optimization algorithm
- real world
- bayesian networks
- real life
- efficient optimization
- virtual world
- robotic control
- optimal control
- optimization methods
- maximum likelihood
- learning process
- global optimization
- function approximation
- augmented reality
- posterior distribution
- optimization model
- action selection
- temporal difference
- simulation environment
- monte carlo sampling
- neural network