知镜 Cognispect · 医学文献与研究工作台

Generalized linear multiple kernel learning for high-dimensional image data

Shi Q, Kang J, Li Y.
Stat Probab Lett · 2026-08-21 · PMID 42677209

文献摘要

We propose GLIMARK, a sparse generalized linear multiple-kernel framework for high-dimensional structured predictors. It extends additive kernel learning to exponential-family outcomes, supports multi-view and hierarchical feature representations, improves scalability via Nyström approximation, and yields interpretable feature-kernel and representer selection.

在知镜精读原文来源DOI 原文

知镜提供医学文献检索、阅读笔记、人体图谱、统计绘图与研究复现工具。