HAL-MLE Log-Splines Density Estimation (Part I: Univariate)
Yilong Hou, Zhengpu Zhao, Yi Li, Mark van der Laan · 2026
Read HAL-MLE Log-Splines Density Estimation (Part I: Univariate) by Yilong Hou, Zhengpu Zhao, Yi Li, Mark van der Laan (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-2602-16259.txt.
Description
Yilong Hou and coauthors study flexible maximum likelihood estimation of densities under a TV-type penalty, known as sectional variation norm. TV regularization has a long history in regression and density work. Revisiting the task through the Highly Adaptive Lasso framework, they formulate a HAL-based estimator using a log-spline link function. In the univariate setting, the bounded section assumption underlying HAL coincides with the classical bounded TV rule.
That equivalence directly connects HAL-MLE to existing TV penalized approaches. Three new theoretical results follow. The univariate HAL-MLE is linear in the limit. It admits pointwise limit normality. It achieves uniform convergence at a stated rate up to logarithmic factors for smoothness order at least one. Those results extend prior work that guaranteed only uniform consistency without rates. The paper aims to unify TV density methods.
Catalog record
| CREATOR | Yilong Hou, Zhengpu Zhao, Yi Li, Mark van der Laan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 154 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-16259