Situated GAIL: Multitask imitation using task-conditioned adversarial inverse reinforcement learning.
Kyoichiro KobayashiTakato HoriiRyo IwakiYukie NagaiMinoru AsadaPublished in: CoRR (2019)
Keyphrases
- multi task
- inverse reinforcement learning
- partially observable environments
- multi task learning
- multitask learning
- preference elicitation
- learning tasks
- reinforcement learning
- reward function
- multi class
- learning problems
- gaussian processes
- transfer learning
- temporal difference
- feature selection
- multi agent
- support vector machine
- information gain
- maximum margin
- data mining
- learning experience
- state space
- active learning