| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 10326513 | Neural Networks | 2008 | 4 Pages |
Abstract
LAGO is an efficient kernel algorithm designed specifically for the rare target detection problem. However, unlike other kernel algorithms, LAGO cannot be easily used with many domain-specific kernels. We solve this problem by first providing a unified framework for LAGO and clarifying its basic principle, and then applying that principle on the unit sphere instead of in the Euclidean space.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Alexandra Laflamme-Sanders, Mu Zhu,
