On Time-Indexing as Inductive Bias in Deep RL for Sequential Manipulation Tasks.
M. Nomaan QureshiBen EisnerDavid HeldPublished in: CoRR (2024)
Keyphrases
- manipulation tasks
- inductive bias
- machine learning
- reinforcement learning
- robot navigation
- inductive learning
- motion planning
- pac learning
- human robot interaction
- robotic systems
- hypothesis space
- supervised learning
- prior knowledge
- domain knowledge
- multitask learning
- learning algorithm
- humanoid robot
- domain theory
- human activities
- real robot
- training examples
- service robots
- vision system
- end effector
- training set
- previously learned
- robotic arm
- degrees of freedom
- state space
- model free
- active learning
- data sets