Gaussian Process Tilted Nonparametric Density Estimation using Fisher Divergence Score Matching
John Paisley, Wei Zhang, Brian Barr · 2025
Read Gaussian Process Tilted Nonparametric Density Estimation using Fisher Divergence Score Matching by John Paisley, Wei Zhang, Brian Barr (2025) 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-2504-03485.txt.
Description
Estimating how likely each outcome is, from data alone, lies at the heart of much statistics. John Paisley, Wei Zhang, and Brian Barr propose a flexible estimator in a 2025 paper. Their density starts from a plain bell curve and tilts it with a learned wavy factor. The tilt comes from a Gaussian process, a flexible curve prior used across machine learning.
Three learning methods are derived in closed form from a score-matching rule. A Fourier trick turns the wavy part into a form like a one-layer network, which makes the math solvable. One method is new: it uses a variational route to mimic a full belief update. Tests compare the three methods with a standard baseline on several data sets.
Catalog record
| CREATOR | John Paisley, Wei Zhang, Brian Barr |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-03485