The MIDAS Touch: Accurate and Scalable Missing-Data Imputation with Deep Learning

Lall, Ranjit; Robinson, Thomas

Abstract

Principled methods for analyzing missing values, based chiefly on multiple imputation, have become increasingly popular yet can struggle to handle the kinds of large and complex data that are also becoming common. We propose an accurate, fast, and scalable approach to multiple imputation, which we call MIDAS (Multiple Imputation with Denoising Autoencoders). MIDAS employs a class of unsupervised neural networks known as denoising autoencoders, which are designed to reduce dimensionality by corrupting and attempting to reconstruct a subset of data. We repurpose denoising autoencoders for multiple imputation by treating missing values as an additional portion of corrupted data and drawing imputations from a model trained to minimize the reconstruction error on the originally observed portion. Systematic tests on simulated as well as real social science data, together with an applied example involving a large-scale electoral survey, illustrate MIDAS’s accuracy and efficiency across a range of settings. We provide open-source software for implementing MIDAS.

Citation

Lall, R., & Robinson, T. (2022). The MIDAS Touch: Accurate and Scalable Missing-Data Imputation with Deep Learning. Political Analysis, 30(2), 179-196. https://doi.org/10.1017/pan.2020.49

• Journal Article Type: Article
• Acceptance Date: Oct 8, 2020
• Online Publication Date: Feb 26, 2021
• Publication Date: 2022-04
• Deposit Date: Dec 14, 2020
• Publicly Available Date: Dec 14, 2020
• Journal: Political Analysis
• Print ISSN: 1047-1987
• Electronic ISSN: 1476-4989
• Publisher: Political Methodology Section
• Peer Reviewed: Peer Reviewed
• Volume: 30
• Issue: 2
• Pages: 179-196
• DOI: https://doi.org/10.1017/pan.2020.49
• Public URL: https://durham-repository.worktribe.com/output/1283564