Trial-and-error learning of repulsors for humanoid QP-based whole-body control.
Jonathan SpitzKarim BouyarmaneSerena IvaldiJean-Baptiste MouretPublished in: Humanoids (2017)
Keyphrases
- learning process
- learning algorithm
- learning problems
- learning systems
- supervised learning
- linear programming
- control rules
- motor control
- real time
- active learning
- genetic algorithm
- prior knowledge
- control system
- mobile devices
- multi agent
- online learning
- multi modal
- unsupervised learning
- objective function
- linear program
- decision trees
- humanoid robot
- adaptive control
- robot control
- sensory motor
- machine learning