Abstract
Random splitting can yield non-independent train--test subsets when a dataset contains related samples, as is common in certain applications such as biochemical studies. This leads to overly optimistic generalization estimates. Here, we introduce ReLaG, a modality-agnostic framework that models sample relatedness through a hierarchical latent-variable process and infers groups of related samples using proximity graphs and community detection to produce independent train--test subsets. Across molecular and protein datasets, ReLaG matches existing relation-aware methods while scaling substantially better, enabling splits at previously impractical dataset sizes. We further introduce a label-free procedure that adapts the splitting resolution to production data, aligning evaluation with the intended deployment setting. ReLaG's inferred groups provide a cheap estimate of effective dataset size, enabling diversity-aware dataset scaling. ReLaG is open source and can be installed with pip install relag.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Lavertu, A., Cote, J., Gobeil, S., Corbeil, J., Premont-Schwarz, I., & Germain, P. (2026). ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations. https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations
MLA 9
Lavertu, Anthony, et al. "ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations." https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations.
Chicago (author–date)
Lavertu, Anthony, Jacob Cote, Sophie Gobeil, Jacques Corbeil, Isabeau Premont-Schwarz, and Pascal Germain. 2026. "ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations." https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations.
Harvard
Lavertu, A., Cote, J., Gobeil, S., Corbeil, J., Premont-Schwarz, I. and Germain, P. (2026) 'ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations', Available at: https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations.
Vancouver
Lavertu A, Cote J, Gobeil S, Corbeil J, Premont-Schwarz I, Germain P. ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations. https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations
IEEE
A. Lavertu, J. Cote, S. Gobeil, J. Corbeil, I. Premont-Schwarz, and P. Germain, "ReLaG: A Scalable Framework Generalizing Random Splits to Data with Latent Relations," https://omanscience.com/en/articles/relag-a-scalable-framework-generalizing-random-splits-to-data-with-latent-relations.