MP3: Movement Primitive-Based (Re-)Planning Policy.
Fabian OttoHongyi ZhouOnur CelikGe LiRudolf LioutikovGerhard NeumannPublished in: CoRR (2023)
Keyphrases
- action selection
- planning problems
- heuristic search
- optimal policy
- partially observable
- reinforcement learning problems
- neural network
- ai planning
- goal oriented
- blocks world
- markov decision problems
- decision support
- supply chain
- domain independent
- path planning
- state space
- reinforcement learning
- asymptotically optimal
- partially observable markov decision processes
- plan generation
- data sets