On model-based clustering with entropic optimal transport
Gonzalo Mena · 2026
Read On model-based clustering with entropic optimal transport by Gonzalo Mena (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-2605-03240.txt.
Introduction
Gonzalo Mena develops a new loss for model-based clustering. It is based on entropic optimal transport. Likelihood clustering is principled, but its surface is nonconvex. The EM algorithm can get trapped at poor points, so users restart it many times and hope one run lands well.
The transport loss shares the same global optimum as the likelihood. Yet its landscape is far better behaved, so many spurious traps vanish. A Sinkhorn-EM optimizer converges at a rate matching EM. In many experiments and two real uses, the new loss beat likelihood methods. The uses were microscopy image segmentation and spatial transcriptomics clustering.
Catalog record
| CREATOR | Gonzalo Mena |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 164 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03240