Improving Patient Subtyping on Longitudinal Data using Representations from Mamba-based Architecture CCS: Computing methodologies Learning latent representations
Md Mozaharul Mottalib, Rahmatollah Beheshti · 2026
Read Improving Patient Subtyping on Longitudinal Data using Representations from Mamba-based Architecture CCS: Computing methodologies Learning latent representations by Md Mozaharul Mottalib, Rahmatollah Beheshti (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-2606-28623.txt.
Description
Md Mozaharul Mottalib and Rahmatollah Beheshti sort patients into meaningful subgroups from their health records. Subtyping promises precision medicine, but records are irregular, visits unevenly spaced, and the time patterns that define a subtype hide inside long, messy sequences that defeat plain clustering.
Their self-supervised model builds on Mamba, a sequence model built for long records, and learns summaries of each patient's history without labels. The team tests the summaries two ways: predicting known labels, and clustering patients into subtypes from the learned geometry. Across public and private real-world datasets, the design choices behind their model deliver better prediction than strong baselines. Head-to-head comparisons of clustering methods show the summaries carry useful subtyping signal. That is a step toward grouping patients by how their conditions unfold.
Catalog record
| CREATOR | Md Mozaharul Mottalib, Rahmatollah Beheshti |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.28623