Cafe: Improved Federated Data Imputation by Leveraging Missing Data Heterogeneity
IEEE Transactions on Knowledge and Data Engineering · 2025-05 · 37 (5) · 2266-2281
DOI: 10.1109/tkde.2025.3537403 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Knowledge and Data Engineering, 2025-05. Verfasst von Sitao Min und 3 weiteren Autor:innen. Beteiligte Einrichtung: Rutgers University, Newark, NJ, USA. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 79 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NLM NIH HHS, NIGMS NIH HHS, NIH.
Abstract (Originalsprache)
Federated learning (FL), a decentralized machine learning approach, offers great performance while alleviating autonomy and confidentiality concerns. Despite FL's popularity, how to deal with missing values in a federated manner is not well understood. In this work, we initiate a study of federated imputation of missing values, particularly in complex scenarios, where missing data heterogeneity exists and the state-of-the-art (SOTA) approaches for federated imputation suffer from significant loss in imputation quality. We propose Cafe, a personalized FL approach for missing data imputation. Cafe is inspired from the observation that heterogeneity can induce differences in observable and missing data distribution across clients, and that these differences can be leveraged to improve the imputation quality. Cafe computes personalized weights that are automatically calibrated for the level of heterogeneity, which can remain unknown, to develop personalized imputation models for each client. An extensive empirical evaluation over a variety of settings demonstrates that Cafe matches the performance of SOTA baselines in homogeneous settings while significantly outperforming the baselines in heterogeneous settings.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Rutgers University, Newark, NJ, USA.
- Hofstra University, Long Island, NY, USA.
Förderung
- NLM NIH HHS
- NIGMS NIH HHS
- NIH
Zitieren
APA
Min, S., Asif, H., Wang, X., & Vaidya, J. (2025). Cafe: Improved Federated Data Imputation by Leveraging Missing Data Heterogeneity. IEEE Transactions on Knowledge and Data Engineering, 37(5), 2266-2281. https://doi.org/10.1109/tkde.2025.3537403
Vancouver
Min S, Asif H, Wang X, Vaidya J. Cafe: Improved Federated Data Imputation by Leveraging Missing Data Heterogeneity. IEEE Transactions on Knowledge and Data Engineering. 2025;37(5):2266-2281. doi:10.1109/tkde.2025.3537403
BibTeX
@article{Min2025,
author = {Min, Sitao and Asif, Hafiz and Wang, Xinyue and Vaidya, Jaideep},
title = {Cafe: Improved Federated Data Imputation by Leveraging Missing Data Heterogeneity},
journal = {IEEE Transactions on Knowledge and Data Engineering},
year = {2025},
volume = {37},
number = {5},
pages = {2266--2281},
doi = {10.1109/tkde.2025.3537403}
}
EUVIMED