I’m a PhD candidate in Machine Learning working on Unsupervised Learning, Clustering, Pattern Recognition, and Natural Language Processing.
My research focuses on clustering methods, validation, and confidence estimation, with applications to structured and textual data.
Affiliated with the Archimedes Research Unit, Athena Research Center.
-
CAKE: Confidence in Assignments via K-partition Ensembles
A framework for estimating confidence in individual clustering assignments by aggregating evidence across multiple partitions. Published in Machine Learning with Applications (2026).
Paper · arXiv · Code · PyPI -
Composite Silhouette: A Subsampling-based Aggregation Strategy
An internal validation criterion for cluster-count selection that combines micro- and macro-averaged Silhouette information across repeated subsamples. Presented at ECML PKDD 2026 and published in the Springer Research Track proceedings.
Paper · arXiv · Code · PyPI -
K-Sil: Silhouette-Driven Instance-Weighted
$k$ -means
A silhouette-driven extension of$k$ -means that weights instances during centroid updates according to their clustering quality.
Preprint · Code · PyPI
Python packages for clustering, validation, confidence estimation, and statistical analysis are available on my PyPI profile, including cake-ensemble, compsil, k-silhouette, sil-score, intclustval, extclustval, and confinterval.
For publications, research, software, NLP work, and other projects: