Reinforcement learning algorithms that assimilate and accommodate skills with multiple tasks.
Paolo TommasinoDaniele CaligioreMarco MirolliGianluca BaldassarrePublished in: ICDL-EPIROB (2012)
Keyphrases
- reinforcement learning algorithms
- multiple tasks
- reinforcement learning
- multi task learning
- state space
- markov decision processes
- model free
- learning algorithm
- multi task
- temporal difference
- reinforcement learning methods
- reward function
- function approximation
- learning problems
- high order
- dynamic environments
- higher order
- multi agent
- feature selection