Disease Phenotyping and Subtyping

We use clustering methods to model multimodal, multivariate, and unevenly-sampled medical time-series data, aiming to uncover hidden disease subtypes and patient cohorts. This enables precise diagnostics, risk stratification, and targeted treatment in complex and heterogeneous conditions.

Publications

Vukosavljević, Katarina, et al. “CAMELOT++: Enhancing Patient Phenotype Discovery through Time-Series Clustering and Survival Analysis.” Women in Machine Learning (WiML) Workshop at NeurIPS, 2024. poster
Aguiar, Henrique, et al. "Learning of cluster-based feature importance for electronic health record time-series." International conference on machine learning. PMLR, 2022. paper