Differential Private Temporal Difference Learning (DPTD)

This is the official implementation for Differentially Private Temporal Difference Learning with Stochastic Nonconvex-Strongly-Concave Optimization (WSDM 2023).

Canzhe Zhao, Yanjie Ze, Jing Dong, Baoxiang Wang, Shuai Li

Instructions

1. Run each .sh file to get the results, e.g., run DPTD in Acrobot env.:

   cd /run/acro
   chmod u+x run_dptd.sh
   ./run_dptd.sh