Identifiable Generative Models for Missing Not at Random Data Imputation

2021/10/27 by Chao Ma, Cheng Zhang, Ma, Chao +1 · 3 citations
Computer Science · Mathematics · #Bayesian Methods and Mixture Models #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Algorithms #Privacy-Preserving Technologies in Data #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.2110.14708

arxiv created 2021/10/27 · openalex publication_date 2021/10/27 · arxiv updated 2021/10/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Real-world datasets often have missing values associated with complex generative processes, where the cause of the missingness may not be fully observed. This is known as missing not at random (MNAR) data. However, many imputation methods do not take into account the missingness mechanism, resulting in biased imputation values when MNAR data is present. Although there are a few methods that have considered the MNAR scenario, their model's identifiability under MNAR is generally not guaranteed. That is, model parameters can not be uniquely determined even with infinite data samples, hence the imputation results given by such models can still be biased. This issue is especially overlooked by many modern deep generative models. In this work, we fill in this gap by systematically analyzing the identifiability of generative models under MNAR. Furthermore, we propose a practical deep generative model which can provide identifiability guarantees under mild assumptions, for a wide range of MNAR mechanisms. Our method demonstrates a clear advantage for tasks on both synthetic data and multiple real-world scenarios with MNAR data.

Citations

• Estimation of Regression Coefficients When Some Regressors are not Always Observed
• Inference and missing data
• Multiple imputation using chained equations: Issues and guidance for practice
• Sample Selection Bias as a Specification Error
• EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAE
• Identification, Doubly Robust Estimation, and Semiparametric Efficiency Theory of Nonignorable Missing Data With a Shadow Variable
• not-MIWAE: Deep Generative Modelling with Missing not at Random Data

Cited by

• Learning-To-Measure: In-Context Active Feature Acquisition
• Deep Generative Models in Condition and Structural Health Monitoring: Opportunities, Limitations and Future Outlook
• Missing Data in Signal Processing and Machine Learning: Models, Methods and Modern Approaches

Related

• Deep Generative Imputation Model for Missing Not At Random Data
• Improving Missing Data Imputation with Deep Generative Models
• Generative Imputation and Stochastic Prediction
• Missing Value Imputation Based on Deep Generative Models
• Missingness Augmentation: A General Approach for Improving Generative Imputation Models