# No imputation without representation

arXiv: https://arxiv.org/abs/2206.14254 · HTML: https://mlsanity.com/paper/2206.14254/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 89 references that are in the map.
- First posted on arXiv 28 Jun 2022.

## Builds on (most cited first, 6 of 6)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [What's a good imputation to predict with missing values?](https://mlsanity.com/paper/2106.00311/) (2021)
- [On the consistency of supervised learning with missing values](https://mlsanity.com/paper/1902.06931/) (2019)
- [Processing of missing data by neural networks](https://mlsanity.com/paper/1805.07405/) (2018)
- [Benchmarking missing-values approaches for predictive models on health databases](https://mlsanity.com/paper/2202.10580/) (2022)
- [Trajectories, bifurcations and pseudotime in large clinical datasets: applications to myocardial infarction and diabetes data](https://mlsanity.com/paper/2007.03788/) (2020)

## Built on by (most cited first, 2 of 2)

- [ITI-IQA: a Toolbox for Heterogeneous Univariate and Multivariate Missing Data Imputation Quality Assessment](https://mlsanity.com/paper/2407.11767/) (2024)
- [Polar Encoding: A Simple Baseline Approach for Classification with Missing Values](https://mlsanity.com/paper/2210.01905/) (2022)

## Abstract (arXiv)

By filling in missing values in datasets, imputation allows these datasets to be used with algorithms that cannot handle missing values by themselves. However, missing values may in principle contribute useful information that is lost through imputation. The missing-indicator approach can be used in combination with imputation to instead represent this information as a part of the dataset. There are several theoretical considerations why missing-indicators may or may not be beneficial, but there has not been any large-scale practical experiment on real-life datasets to test this question for machine learning predictions. We perform this experiment for three imputation strategies and a range of different classification algorithms, on the basis of twenty real-life datasets. In a follow-up experiment, we determine attribute-specific missingness thresholds for each classifier above which missing-indicators are more likely than not to increase classification performance. And in a second follow-up experiment, we evaluate numerical imputation of one-hot encoded categorical attributes. We reach the following conclusions. Firstly, missing-indicators generally increase classification performance. Secondly, with missing-indicators, nearest neighbour and iterative imputation do not lead to better performance than simple mean/mode imputation. Thirdly, for decision trees, pruning is necessary to prevent overfitting. Fourthly, the thresholds above which missing-indicators are more likely than not to improve performance are lower for categorical attributes than for numerical attributes. Lastly, mean imputation of numerical attributes preserves some of the information from missing values. Consequently, when not using missing-indicators it can be advantageous to apply mean imputation to one-hot encoded categorical attributes instead of mode imputation.