Eigen-guided transformer: A data-driven approach for chronic kidney disease forecasting
Fahman Saeed, Sultan Aldera · 2026
Read Eigen-guided transformer: A data-driven approach for chronic kidney disease forecasting by Fahman Saeed, Sultan Aldera (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/plos-pone-0348886.txt.
Description
Fahman Saeed and Sultan Aldera present a deep learning model for forecasting creatinine, a key marker of kidney function, in hospital patient records. Predicting chronic kidney disease early is hard because each patient's course differs and risk factors tangle together. Their model is a transformer, the architecture behind modern language AI, reshaped by statistics drawn from the data itself.
Eigenvalue analysis of how clinical features correlate set the model's width, and the same math seeded its starting weights so it begins with real feature relationships built in. Tested on daily sequences from a large critical care database, it predicted creatinine with low error. The design also answers the black-box complaint: its structure is derived openly from the data rather than chosen by guesswork.
Catalog record
| CREATOR | Fahman Saeed, Sultan Aldera |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57637 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |