To discretize continually: Mean shift interacting particle systems for Bayesian inference
Ayoub Belhadji, Daniel Sharp, Youssef Marzouk · 2026
Read To discretize continually: Mean shift interacting particle systems for Bayesian inference by Ayoub Belhadji, Daniel Sharp, Youssef Marzouk (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-2605-14142.txt.
Description
Belhadji, Sharp and Marzouk introduce mean shift particle systems for Bayesian inference. Adding up a distribution known only through its unnormalized density is a central Bayesian task. Their methods guess such means with a few weighted samples, a quadrature rule, built by a particle system that shrinks mean discrepancy to the target. The approach extends the old mean shift algorithm and recent optimal quantization methods to smooth distributions. Its motion ignores the unknown normalizing constant and comes in gradient-free and gradient-informed versions.
Catalog record
| CREATOR | Ayoub Belhadji, Daniel Sharp, Youssef Marzouk |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60294 |
| LICENSE | http://creativecommons.org/licenses/by-sa/4.0/ |
| RIGHTS | CC BY SA |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14142