Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Review for Handling Missing Data with special missing mechanism

Youran Zhou, Sunil Aryal, Mohamed Reda Bouadjenek

TL;DR

The paper addresses the challenge of missing data in tabular datasets by foregrounding special missing mechanisms, notably MAR and MNAR, which are less understood than MCAR. It surveys a broad spectrum of imputation methods, ranging from traditional statistical and machine-learning approaches to advanced deep learning and optimization-based techniques, and it highlights methodological gaps in MAR/MNAR handling and data-generation practices. The authors categorize existing methods into deletion, imputation, and representation learning, and they provide guidance on generating MAR/MNAR data for robust evaluation. The work aims to standardize benchmarking, promote deep-learning-driven imputation strategies, and offer actionable directions for future research to improve imputation accuracy and downstream utility in real-world scenarios.

Abstract

Missing data poses a significant challenge in data science, affecting decision-making processes and outcomes. Understanding what missing data is, how it occurs, and why it is crucial to handle it appropriately is paramount when working with real-world data, especially in tabular data, one of the most commonly used data types in the real world. Three missing mechanisms are defined in the literature: Missing Completely At Random (MCAR), Missing At Random (MAR), and Missing Not At Random (MNAR), each presenting unique challenges in imputation. Most existing work are focused on MCAR that is relatively easy to handle. The special missing mechanisms of MNAR and MAR are less explored and understood. This article reviews existing literature on handling missing values. It compares and contrasts existing methods in terms of their ability to handle different missing mechanisms and data types. It identifies research gap in the existing literature and lays out potential directions for future research in the field. The information in this review will help data analysts and researchers to adopt and promote good practices for handling missing data in real-world problems.

Review for Handling Missing Data with special missing mechanism

TL;DR

The paper addresses the challenge of missing data in tabular datasets by foregrounding special missing mechanisms, notably MAR and MNAR, which are less understood than MCAR. It surveys a broad spectrum of imputation methods, ranging from traditional statistical and machine-learning approaches to advanced deep learning and optimization-based techniques, and it highlights methodological gaps in MAR/MNAR handling and data-generation practices. The authors categorize existing methods into deletion, imputation, and representation learning, and they provide guidance on generating MAR/MNAR data for robust evaluation. The work aims to standardize benchmarking, promote deep-learning-driven imputation strategies, and offer actionable directions for future research to improve imputation accuracy and downstream utility in real-world scenarios.

Abstract

Missing data poses a significant challenge in data science, affecting decision-making processes and outcomes. Understanding what missing data is, how it occurs, and why it is crucial to handle it appropriately is paramount when working with real-world data, especially in tabular data, one of the most commonly used data types in the real world. Three missing mechanisms are defined in the literature: Missing Completely At Random (MCAR), Missing At Random (MAR), and Missing Not At Random (MNAR), each presenting unique challenges in imputation. Most existing work are focused on MCAR that is relatively easy to handle. The special missing mechanisms of MNAR and MAR are less explored and understood. This article reviews existing literature on handling missing values. It compares and contrasts existing methods in terms of their ability to handle different missing mechanisms and data types. It identifies research gap in the existing literature and lays out potential directions for future research in the field. The information in this review will help data analysts and researchers to adopt and promote good practices for handling missing data in real-world problems.
Paper Structure (53 sections, 8 equations, 11 figures, 5 tables)

This paper contains 53 sections, 8 equations, 11 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Contribution
  3. Background and Preliminary
  4. Missing Rate
  5. Missing Pattern
  6. Missing Mechanisms
  7. Missing Completely at Random (MCAR)
  8. Missing At Random (MAR)
  9. Missing Not At Random(MNAR)
  10. Related Work
  11. Contribution
  12. Existing Methods of Handling Missing Data
  13. Taxonomy
  14. Deletion for Handling Missing Data
  15. Listwise Deletion
  16. ...and 38 more sections

Figures (11)

  • Figure 1: Number of articles on handling missing data retrieved by Scopus based on keyword searches with and without the special mechanisms keywords of MAR or MNAR
  • Figure 2: Comparison of missing rates in different features. The left plot illustrates evenly distributed missing data, where each feature has similar missing rates. On the right plot, unevenly distributed missing data is shown, with varying missing rates across features. This visual representation highlights the diversity in missing data patterns within the dataset.
  • Figure 3: Missing data patterns in multivariate data. Blue is observed $X^o$, green is missing part $X^m$. The missing pattern from left to right are: Univariate, Monotone, Filematching and General
  • Figure 4: Method Taxonomy for Handling Missing Data
  • Figure 5: Keyword search from Scopus database for MNAR and MAR Data Imputation (All data types) Method by Year(2000-2023)
  • ...and 6 more figures