Attention-Based Deep Learning for Early Parkinson’s Disease Detection with Tabular Biomedical Data
Olamide S. Oseni, Ibraheem O. Obanla, Toheeb A. Jimoh · 2026
Read Attention-Based Deep Learning for Early Parkinson’s Disease Detection with Tabular Biomedical Data by Olamide S. Oseni, Ibraheem O. Obanla, Toheeb A. Jimoh (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-07933.txt.
Description
Parkinson's disease is hardest to catch early, when symptoms are subtle and biomedical signals interact in complex ways. Olamide Samuel Oseni and two co-authors compare attention-based deep learning models against traditional methods for early detection from tabular biomedical data, using voice measurements from patients and healthy controls.
Four models are tested, including a multilayer perceptron, gradient boosting, TabNet, and SAINT. SAINT leads on every metric, reaching weighted precision of 0.98, recall of 0.97, and an F1-score of 0.97, with a Matthews correlation coefficient near 0.999. The study suggests attention mechanisms can capture the feature interactions that simpler models miss, offering clinicians a promising screening aid.
Catalog record
| CREATOR | Olamide S. Oseni, Ibraheem O. Obanla, Toheeb A. Jimoh |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-07933