Unsupervised Imputation of Non-ignorably Missing Data Using Importance-Weighted Autoencoders

TL;DR

This paper introduces NIMIWAE, a novel Variational Autoencoder architecture designed to effectively handle both ignorable and non-ignorable missing data patterns in biomedical datasets, improving imputation and analysis.

Contribution

The paper presents NIMIWAE, the first VAE model that flexibly models both missing data mechanisms, enhancing unsupervised learning and imputation in complex biomedical data.

Findings

Outperforms existing imputation methods in simulations

Accurately models non-ignorable missingness patterns

Effective in real-world ICU dataset analysis

Abstract

Deep Learning (DL) methods have dramatically increased in popularity in recent years. While its initial success was demonstrated in the classification and manipulation of image data, there has been significant growth in the application of DL methods to problems in the biomedical sciences. However, the greater prevalence and complexity of missing data in biomedical datasets present significant challenges for DL methods. Here, we provide a formal treatment of missing data in the context of Variational Autoencoders (VAEs), a popular unsupervised DL architecture commonly utilized for dimension reduction, imputation, and learning latent representations of complex data. We propose a new VAE architecture, NIMIWAE, that is one of the first to flexibly account for both ignorable and non-ignorable patterns of missingness in input features at training time. Following training, samples can be drawn from the approximate posterior distribution of the missing data can be used for multiple imputation, facilitating downstream analyses on high dimensional incomplete datasets. We demonstrate through statistical simulation that our method outperforms existing approaches for unsupervised learning tasks and imputation accuracy. We conclude with a case study of an EHR dataset pertaining to 12,000 ICU patients containing a large number of diagnostic measurements and clinical outcomes, where many features are only partially observed.