Skip to content

ywang485/pbcplus2pomdp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

pbcplus2pomdp

Python scripts that translate probabilistic action descriptions in pBC+ (encodede as lpmln programs) to POMDP models. The outputs are .pomdp files that can be used as input to POMDP solver APPL (https://github.com/AdaCompNUS/sarsop).

Usage

Noncompositional generation:

$ python pbcplus2pomdp.py path/to/lpmln/files discount_factor(0-1)

Compositional generation:

$ python pbcplus2pomdp_compositional.py path/to/lpmln/file/with/no/action path/to/lpmln/action/description/files discount_factor(0-1)

For example,

$ python pbcplus2pomdp.py examples/tiger.lpmln 0.95

$ python pbcplus2pomdp_compositional.py examples/dialog/dialog-comp/dialog_noaction.lpmln examples/dialog/dialog-comp/actions 0.95

System Dependencies

This Python script requires the following system to be installed

Examples

The folder "examples" contains lpmln encodings for several example pBC+ action description.

  • tiger.lpmln: the well-known tiger example in POMDP literature

  • two_tiger.lpmln: a variation of the tiger example where there are two tigers and three doors

  • dialog: an example from 1

References

[1] Zhang, Shiqi, Piyush Khandelwal, and Peter Stone. "Dynamically constructed (po) mdps for adaptive robot planning." Thirty-First AAAI Conference on Artificial Intelligence. 2017.

About

A tool that generates POMDP models from pBC+ action descriptions

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages