Preprint · CC-BY
via OpenAlex
Age Predictors Through the Lens of Generalization, Bias Mitigation, and Interpretability: Reflections on Causal Implications
Debdas Paul, Elisa Ferrari, Irene Gravili, Alessandro Cellerino
arXiv (Cornell University) · 2026
Abstract
Chronological age predictors often fail to achieve out-of-distribution (OOD) gen- eralization due to exogenous attributes such as race, gender, or tissue. Learning an invariant representation with respect to those attributes is therefore essential to improve OOD generalization and prevent overly optimistic results. In predic- tive settings, these attributes motivate bias mitigation; in causal analyses, they appear as confounders; and when protected, their suppression leads to fairness. We coherently explore these concepts with theoretical rigor and discuss the scope of an interpretable neural network model based on adversarial representation learning. Using publicly available mouse transcriptomic datasets, we illustrate the behavior of this model relative to conventional machine learning models. We observe that the outcome of this model is consistent with the predictive results of a published study demonstrating the effects of Elamipretide on mouse skeletal and cardiac muscle. We conclude by discussing the limitations of deriving causal interpretation from such purely predictive models.
◌ CITATION ONLY
Full text is not openly licensed for redistribution here. Read it at the source:
Provenance
- Source
- OpenAlex
- DOI
- 10.48550/arxiv.2603.16377
- Canonical
- link ↗
- Fetched
- 2026-11-08 MST
Cite this
APA
Paul, D., Ferrari, E., Gravili, I., & Cellerino, A. (2026). Age Predictors Through the Lens of Generalization, Bias Mitigation, and Interpretability: Reflections on Causal Implications. <em>arXiv (Cornell University)</em>. https://doi.org/10.48550/arxiv.2603.16377
Vancouver
Paul D, Ferrari E, Gravili I, Cellerino A. Age Predictors Through the Lens of Generalization, Bias Mitigation, and Interpretability: Reflections on Causal Implications. arXiv (Cornell University). 2026. doi:10.48550/arxiv.2603.16377.
BibTeX
@unpublished{debdas2026AgePre,
title = {Age Predictors Through the Lens of Generalization, Bias Mitigation, and Interpretability: Reflections on Causal Implications},
author = {Debdas Paul and Elisa Ferrari and Irene Gravili and Alessandro Cellerino},
journal = {arXiv (Cornell University)},
year = {2026},
doi = {10.48550/arxiv.2603.16377},
}
Research neighborhood
References, citing works, and semantically nearest findings. Click a node to open it.
Related findings
ArXiv.org 2026
Preprint · CC-BY
Age Predictors Through the Lens of Generalization, Bias Mitigation, and Interpretability: Reflections on Causal Implications
David M. Brown, M.D. 2012
Open access · US-GOV
A Phase I/II, Randomized, Study for Diabetic Macular Edema Using 0.3mg Ranibizumab Combined With Targeted PRP Monthly for 4 Months,Then PRN vs. 0.3mg Ranibizumab 4 Months Monotherapy, Then as Needed(DME-AntiVEgf) DAVE
Journal of Endocrinology 2018
Open access · OA
Metabolic effects of short-term caloric restriction in mice with reduced insulin gene dosage
The Journal of Maternal-Fetal & Neonatal Medicine 2021
Open access · CC-BY
Proteomic signatures predict preeclampsia in individual cohorts but not across cohorts – implications for clinical biomarker studies
Aging and Disease 2025
Open access · CC-BY
The Ability of Activity and Cognition in Old Mice to Predict Age of Death
2023
Open access · CC-BY