Skip to content
/ lmdp Public

Lexicographic value iteration for LMDPs with slack and conditional preferences.

Notifications You must be signed in to change notification settings

kylewray/lmdp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

63 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

lmdp

Lexicographic Markov Decision Processes (LMDPs) are MOMDPs with state-dependent lexicographic preferences over the reward functions, allowing for slack in optimization. Value iteration for LMDPs solves this problem by applying dynamic programming over the states and rewards in a particular order, yielding one of the solutions which satisfy the slack constraints of the LMDP.

For more information, please see our AAAI 2015 paper:

Wray, Kyle H., Zilberstein, Shlomo, and Mouaddib, Abdel-Illah. "Multi-Objective MDPs with Conditional Lexicographic Reward Preferences." In Proceedings of the Twenty Ninth Conference on Artificial Intelligence (AAAI), Austin, TX, USA, January 2015.

About

Lexicographic value iteration for LMDPs with slack and conditional preferences.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published