Use Case Extensions

The interfaces provided by pomdp_py should support projects in these directions (and more):

1. POMDP for RL (Reinforcement Learning)

   • Learn a PolicyModel (model-free)

   • Learn a TransitionModel and ObservationModel (model-based)

2. Multi-Agent POMDPs

   • Define multiple Agent classes.

3. Object-Oriented POMDPs

   • The Object-Oriented POMDP (OO-POMDP) already has its interfaces implemented in oopomdp.

3. Task Transfer or Transfer Learning

   • Define multiple Agent classes.

3. Planning Algorithm Research

   • Use existing POMCP or POUCT as baslines.