Leveraging variational autoencoders for multiple data imputation
Missing data persists as a major barrier to data analysis across numerous applications. Recently, deep generative models have been used for imputation of missing data, motivated by their ability to capture highly non-linear and complex relationships in the data. In this work, we investigate the ability of deep models, namely variational autoencoders (VAEs), to account for uncertainty in missing data through multiple imputation strategies. We find that VAEs provide poor empirical coverage of missing data, with underestimation and overconfident imputations, particularly for more extreme missing data values. To overcome this, we employ $\beta$-VAEs, which viewed from a generalized Bayes framework, provide robustness to model misspecification. Assigning a good value of $\beta$ is critical for uncertainty calibration and we demonstrate how this can be achieved using cross-validation. In downstream tasks, we show how multiple imputation with $\beta$-VAEs can avoid false discoveries that arise as artefacts of imputation.
Code (1)
Tasks
ImputationSimilar Papers 제목 키워드 기반
Multiple Imputation for Biomedical Data using Monte Carlo Dropout Autoencoders
Due to complex experimental settings, missing values are common in biomedical data. To handle this issue, many methods have been proposed, from ignoring incomplete instances to various data imputation approaches. With th…
ImputationMissing ValuesArbitrarily-conditioned Data Imputation
In this paper, we propose an arbitrarily-conditioned data imputation framework built upon variational autoencoders and normalizing flows. The proposed model is capable of mapping any partial data to a multi-modal latent …
ImputationLearning from Partially-Observed Multimodal Data with Variational Autoencoders
Learning from only partially-observed data for imputation has been an active research area. Despite promising progress on unimodal data imputation (e.g., image in-painting), models designed for multimodal data imputation…
ImputationMissing Data Imputation using Neural Cellular Automata
When working with tabular data, missingness is always one of the most painful problems. Throughout many years, researchers have continuously explored better and better ways to impute missing data. Recently, with the rapi…
Continuous-time Autoencoders for Regular and Irregular Time Series Imputation
Time series imputation is one of the most fundamental tasks for time series. Real-world time series datasets are frequently incomplete (or irregular with missing observations), in which case imputation is strongly requir…
ImputationIrregular Time SeriesTime Series