Code for Paper "Robust Chance-constrained Policy Synthesis and Verification via Moment-Based Distributional Analysis"
This repository contains implementations and experimental code for robust chance-constrained policy synthesis using moment-based distributional analysis.
-
mdp.py: the environment implementation adapted from a PRISM case study. This file defines the finite MDP used in our experiments. -
policy_synthesis.ipynb: the notebook demo comparing (1) risk-neutral value iteration (baseline) and (2) robust chance-constrained policy synthesis using moment-based distributional analysis.
The notebook provides an end-to-end, runnable example on the betting-game environment adapted from a PRISM case study.