Integrated Imputation-Classification for Supervised Learning with Missing Data

arXiv:2610.04273v1 Announce Type: new Abstract: We study supervised classification problems with missing feature values. Existing approaches often decouple imputation from classification, producing imputations that may be plausible but uninformative for prediction. Instead, we propose the Integrated Imputation and Classification Network (IICN), which jointly trains an imputer and an {(n{+}1)}-classdiscriminator adversarially with a single class supervised classification objective, where the discriminator learns to distinguish among the n true classes and an additional ``imputed" class. We prove that at the global optimum, the imputer and discriminator together implement marginalization over missing coordinates and yield a Bayes-optimal classifier. We evaluate IICN on FashionMNIST, CIFAR-10, and tabular datasets with naturally occurring missingness. IICN outperforms classical impute-then-classify pipelines and recent generative baselines, showing strong robustness and accuracy in challenging settings.

Related

• GLR-MM: Graph-Based Global-Local Reconstruction for Robust Multimodal Chest X-ray and EHR Representation Learning under Missing Modalities
• Impute-EM: Native Mixed-State Diffusion Models for Heterogeneous Data Imputation
• Latent Diffusion for Missing Data
• Probabilistic indirect models for undrained shear strength: addressing significant data missing and variability with advanced imputation and machine learning techniques

Source: arXiv cs.LG | 2026-10-06