grv

Gradient Regularized V-Learning for Dynamic Treatment Regimes

This is a python implementation of the algorithms in the paper "Gradient Regularized V-Learning for Dynamic Treatment Regimes" published in NeurIPS 2020. The goal of this algorithm is to evaluate treatment rules or find the optimal treatment rules from observational data.

Dependencies

Python 3.6 or later and Tensorflow 1.9.0, see requirements.txt.

Examples

python train_grvb.py
python train_grvs.py