Identifiable Deep Latent Variable Models for MNAR Data

• URL: http://arxiv.org/abs/2603.24771v2
• Date: Fri, 27 Mar 2026 02:53:11 GMT
• Title: Identifiable Deep Latent Variable Models for MNAR Data
• Abstract summary: Missing data is a ubiquitous challenge in data analysis, often leading to biased and inaccurate results.<n>Traditional imputation methods usually assume that the missingness mechanism is missing-at-random (MAR), where the missingness is independent of the missing values themselves.<n>This paper proposes a novel framework based on deep latent variable models for missing-not-at-random (MNAR) data.
• Score: 13.739530975229863
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Missing data is a ubiquitous challenge in data analysis, often leading to biased and inaccurate results. Traditional imputation methods usually assume that the missingness mechanism is missing-at-random (MAR), where the missingness is independent of the missing values themselves. This assumption is frequently violated in real-world scenarios, prompted by recent advances in imputation methods using deep learning to address this challenge. However, these methods neglect the crucial issue of nonparametric identifiability in missing-not-at-random (MNAR) data, which can lead to biased and unreliable results. This paper seeks to bridge this gap by proposing a novel framework based on deep latent variable models for MNAR data. Building on the assumption of conditional no self-censoring given latent variables, we establish the identifiability of the data distribution. This crucial theoretical result guarantees the feasibility of our approach. To effectively estimate unknown parameters, we develop an efficient algorithm utilizing importance-weighted autoencoders. We demonstrate, both theoretically and empirically, that our estimation process accurately recovers the ground-truth joint distribution under specific regularity conditions. Extensive simulation studies and real-world data experiments showcase the advantages of our proposed method compared to various classical and state-of-the-art approaches to missing data imputation.

Related papers

• Robust Uncertainty Estimation under Distribution Shift via Difference Reconstruction [2.920726125254978]
  Estimating uncertainty in deep learning models is critical for reliable decision-making in high-stakes applications such as medical imaging.<n>Prior research has established that the difference between an input sample and its reconstructed version can serve as a useful proxy for uncertainty.<n>We propose Difference Reconstruction Uncertainty Estimation (DRUE), a method that mitigates this limitation by reconstructing inputs from two intermediate layers and measuring the discrepancy between their outputs as the uncertainty score.
  arXiv  Detail & Related papers  (2026-01-27T08:23:00Z)
• Revisiting Multivariate Time Series Forecasting with Missing Values [65.30332997607141]
  Missing values are common in real-world time series.<n>Current approaches have developed an imputation-then-prediction framework that uses imputation modules to fill in missing values, followed by forecasting on the imputed data.<n>This framework overlooks a critical issue: there is no ground truth for the missing values, making the imputation process susceptible to errors that can degrade prediction accuracy.<n>We introduce Consistency-Regularized Information Bottleneck (CRIB), a novel framework built on the Information Bottleneck principle.
  arXiv  Detail & Related papers  (2025-09-27T20:57:48Z)
• Recursive Equations For Imputation Of Missing Not At Random Data With Sparse Pattern Support [8.863778901027061]
  A common approach for handling missing values in data analysis pipelines is multiple imputation via software packages.<n>We develop a new characterization for the full data law in graphical models of missing data.<n>We show MISPR obtains comparable results to MICE when data are MAR, and superior, less biased results when data are MNAR.
  arXiv  Detail & Related papers  (2025-07-21T23:18:36Z)
• Robust Molecular Property Prediction via Densifying Scarce Labeled Data [53.24886143129006]
  In drug discovery, compounds most critical for advancing research often lie beyond the training set.<n>We propose a novel bilevel optimization approach that leverages unlabeled data to interpolate between in-distribution (ID) and out-of-distribution (OOD) data.
  arXiv  Detail & Related papers  (2025-06-13T15:27:40Z)
• DAGnosis: Localized Identification of Data Inconsistencies using Structures [73.39285449012255]
  Identification and appropriate handling of inconsistencies in data at deployment time is crucial to reliably use machine learning models. We use directed acyclic graphs (DAGs) to encode the training set's features probability distribution and independencies as a structure. Our method, called DAGnosis, leverages these structural interactions to bring valuable and insightful data-centric conclusions.
  arXiv  Detail & Related papers  (2024-02-26T11:29:16Z)
• Sufficient Identification Conditions and Semiparametric Estimation under Missing Not at Random Mechanisms [4.211128681972148]
  Conducting valid statistical analyses is challenging in the presence of missing-not-at-random (MNAR) data. We consider a MNAR model that generalizes several prior popular MNAR models in two ways. We propose methods for testing the independence restrictions encoded in such models using odds ratio as our parameter of interest.
  arXiv  Detail & Related papers  (2023-06-10T13:46:16Z)
• MissDAG: Causal Discovery in the Presence of Missing Data with Continuous Additive Noise Models [78.72682320019737]
  We develop a general method, which we call MissDAG, to perform causal discovery from data with incomplete observations. MissDAG maximizes the expected likelihood of the visible part of observations under the expectation-maximization framework. We demonstrate the flexibility of MissDAG for incorporating various causal discovery algorithms and its efficacy through extensive simulations and real data experiments.
  arXiv  Detail & Related papers  (2022-05-27T09:59:46Z)
• MIRACLE: Causally-Aware Imputation via Learning Missing Data Mechanisms [82.90843777097606]
  We propose a causally-aware imputation algorithm (MIRACLE) for missing data. MIRACLE iteratively refines the imputation of a baseline by simultaneously modeling the missingness generating mechanism. We conduct extensive experiments on synthetic and a variety of publicly available datasets to show that MIRACLE is able to consistently improve imputation.
  arXiv  Detail & Related papers  (2021-11-04T22:38:18Z)
• Identifiable Generative Models for Missing Not at Random Data Imputation [13.790820495804567]
  Many imputation methods do not take into account the missingness mechanism, resulting in biased imputation values when MNAR data is present. In this work, we analyze the identifiability of generative models under MNAR. We propose a practical deep generative model which can provide identifiability guarantees under mild assumptions.
  arXiv  Detail & Related papers  (2021-10-27T18:51:38Z)
• Unlabelled Data Improves Bayesian Uncertainty Calibration under Covariate Shift [100.52588638477862]
  We develop an approximate Bayesian inference scheme based on posterior regularisation. We demonstrate the utility of our method in the context of transferring prognostic models of prostate cancer across globally diverse populations.
  arXiv  Detail & Related papers  (2020-06-26T13:50:19Z)
• Uncertainty-Gated Stochastic Sequential Model for EHR Mortality Prediction [6.170898159041278]
  We present a novel variational recurrent network that estimates the distribution of missing variables, updates hidden states, and predicts the possibility of in-hospital mortality. It is noteworthy that our model can conduct these procedures in a single stream and learn all network parameters jointly in an end-to-end manner.
  arXiv  Detail & Related papers  (2020-03-02T04:41:28Z)
• MissDeepCausal: Causal Inference from Incomplete Data Using Deep Latent Variable Models [14.173184309520453]
  State-of-the-art methods for causal inference don't consider missing values. Missing data require an adapted unconfoundedness hypothesis. Latent confounders whose distribution is learned through variational autoencoders adapted to missing values are considered.
  arXiv  Detail & Related papers  (2020-02-25T12:58:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.