AutoML-Multiverse: An Instability-Aware Framework for Quantifying Analytic Variability in Alzheimer’s Disease Machine-Learning Studies
AutoML-Multiverse systematically quantifies analytic instability in clinical ML.Analysis of ∼20,000 pipelines across ADNI and NACC cohorts.Pipeline choices substantially alter model rankings and biomarker importance.Cross-cohort variability highlights risks of single-dataset studies.Instability-aware evaluation improves robustness of AI-driven research.
Authors
- Robert Leech (ORCID: https://orcid.org/0000-0002-5801-6318)
- D H Wyllie
- Peter Weston
- Maitrei Kohli
- Gonzalo Castro Leal
- Neil P Oxtoby
- James H Cole
Institutions
- King's College London (GB)
- UK Dementia Research Institute (GB)
- The London College (GB)
- National Hospital for Neurology and Neurosurgery (GB)
- University College London (GB)
Publication Details
- Journal
- medRxiv
- Published
- 2026-03-16
- DOI
- https://doi.org/10.64898/2026.03.13.26347938
- Primary Topic
- Dementia and Cognitive Impairment Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00