/GAIN-Pytorch

Pytorch implementation of GAIN for missing data imputation

Primary LanguageJupyter Notebook

Generative Adversarial Imputation Networks (GAIN) Pytorch Implementation

Pytorch implementation of the paper GAIN: Missing Data Imputation using Generative Adversarial Nets by Jinsung Yoon, James Jordon, Mihaela van der Schaar

Reference: J. Yoon, J. Jordon, M. van der Schaar, "GAIN: Missing Data Imputation using Generative Adversarial Nets," International Conference on Machine Learning (ICML), 2018.

This notebook is a Pytorch adaptation of the original Tensorflow code available here: https://github.com/jsyoon0823/GAIN

This repo is tested on Python 3.6 and PyTorch 1.4.0.

The code can be run in either GPU or CPU (using use_gpu flag).

The results are same as the original implementation.

Datasets are taken from the original implementation.