Here's the code base of algorthm that can solve POMDPs, for example, PBVI (Point Based Value Iteration). The project will continuously update.
File Organization
-
runPomdpExample.py: Executable example.
-
src
- pomdpSimulation.py: Simulate the POMDP world (For model evaluation part).
- pbvi.py: Core code of PBVI (Point Based Value Iteration).
- pbviBeliefExpension.py: Belief Expension Method for PBVI.
- tigermodel.py: Tiger Model example.
- tagmodel.py: Tag Model example.
-
Environment
- tigerEnvironment.py: Task Environment of Two-Tiger Problem.
- tagEnvironment.py: Task Environment of Tag Problem.
-
tools: Useful packed tools for solving POMDPs.
- alphaVector.py: Alpha-Vector data structure.
- sampleUtility.py: Sampling method packages.
-
exec: Model evaluation for different conditions (different parameter settings).
-
visualization
- visualizeTiger.py: Visualize tiger model structure.
- visualizeExec: Visualize the evaluation results.