Repository for code release of paper "Robust Variational Autoencoders for Outlier Detection and Repair of Mixed-Type Data" (AISTATS 2020)
Code for paper: "Robust Variational Autoencoders for Outlier Detection and Repair of Mixed-Type Data" AISTATS 2020. Check it here https://arxiv.org/abs/1907.06671. Please consider citing us if you use our code.
Please install ./setup.py in folder ./src in order to use core_models package.
Use Pytorch 1.3.1 at least
python noising_process.py
in separate folders.python noising_process.py
--cuda-on
for GPU. For instance:
sh run_RVAE_CVI.sh
, for our main algorithm.sh run_VAE_l2.sh
, for VAE-L2 baseline.sh run_CondPred.sh
, for NN-based Conditional Predictor (pseudo-likelihood).sh run_baselines.sh
, for assorted baselines in paper.MIT