Low Dimensional State Representation Learning with Reward-shaped Priors.
Nicolò BotteghiRuben ObbinkDaan GeijsMannes PoelBeril SirmaçekChristoph BruneAbeje Y. MershaStefano StramigioliPublished in: CoRR (2020)
Keyphrases
- learning process
- low dimensional
- reinforcement learning
- knowledge acquisition
- learning systems
- multiple representations
- dynamic bayesian networks
- pattern recognition
- high dimensional
- prior knowledge
- mobile robot
- partially observable
- latent variables
- background knowledge
- online learning
- learning algorithm
- multi agent
- image processing
- computer vision