DeR-CFR

Introduction

This repository contains the implementation code for paper:

Learning Decomposed Representations for Treatment Effect Estimation

Anpeng Wu*, Junkun Yuan*, Kun Kuang, Bo Li, Runze Wu, Qiang Zhu, Yueting Zhuang, and Fei Wu

https://arxiv.org/abs/2006.07040

Requirements

Hardware configuration: Ubuntu 16.04.5 LTS operating system with 2 * Intel Xeon E5-2678 v3 CPU, 384GB of RAM, and 4 * GeForce GTX 1080Ti GPU with 44GB of VRAM.

Software configuration: Python with TensorFlow 1.15.0, NumPy 1.17.4, and MatplotLib 3.1.1.

Create Env:

conda env create -f environment.yaml

Instructions

Run generator_data.py scripts to build data, and then Train.py scripts to run experiments on the generated data and save results into "results" directory.

python generator_data.py
python Train.py