EGG-GAE: scalable graph neural networks for tabular data imputation
This addresses scalability issues in graph-based imputation for tabular data, offering a practical solution for domains with large datasets, though it is incremental over prior graph autoencoder methods.
The paper tackles the problem of missing data imputation in tabular datasets by proposing EGG-GAE, a scalable graph autoencoder that automatically infers connectivity and works on mini-batches, resulting in significant improvements in imputation error and downstream accuracy across multiple benchmarks.
Missing data imputation (MDI) is crucial when dealing with tabular datasets across various domains. Autoencoders can be trained to reconstruct missing values, and graph autoencoders (GAE) can additionally consider similar patterns in the dataset when imputing new values for a given instance. However, previously proposed GAEs suffer from scalability issues, requiring the user to define a similarity metric among patterns to build the graph connectivity beforehand. In this paper, we leverage recent progress in latent graph imputation to propose a novel EdGe Generation Graph AutoEncoder (EGG-GAE) for missing data imputation that overcomes these two drawbacks. EGG-GAE works on randomly sampled mini-batches of the input data (hence scaling to larger datasets), and it automatically infers the best connectivity across the mini-batch for each architecture layer. We also experiment with several extensions, including an ensemble strategy for inference and the inclusion of what we call prototype nodes, obtaining significant improvements, both in terms of imputation error and final downstream accuracy, across multiple benchmarks and baselines.