Variational Weighting for Kernel Density Ratios
Abstract
Kernel density estimation (KDE) is integral to a range of generative and discriminative tasks in machine learning. Drawing upon tools from the multidimensional calculus of variations, we derive an optimal weight function that reduces bias in standard kernel density estimates for density ratios, leading to improved estimates of prediction posteriors and information-theoretic measures. In the process, we shed light on some fundamental aspects of density estimation, particularly from the perspective of algorithms that employ KDEs as their main building blocks.
Cite
Text
Yoon et al. "Variational Weighting for Kernel Density Ratios." Neural Information Processing Systems, 2023.Markdown
[Yoon et al. "Variational Weighting for Kernel Density Ratios." Neural Information Processing Systems, 2023.](https://mlanthology.org/neurips/2023/yoon2023neurips-variational/)BibTeX
@inproceedings{yoon2023neurips-variational,
title = {{Variational Weighting for Kernel Density Ratios}},
author = {Yoon, Sangwoong and Park, Frank and Yun, Gunsu and Kim, Iljung and Noh, Yung-Kyun},
booktitle = {Neural Information Processing Systems},
year = {2023},
url = {https://mlanthology.org/neurips/2023/yoon2023neurips-variational/}
}