A Unified View of Localized Kernel Learning

TL;DR

This paper presents a unified framework for localized kernel learning, improving upon existing methods by designing a new algorithm that enhances scalability and performance in kernel-based models.

Contribution

It unifies various localized kernel learning approaches into a single system and introduces a new, more effective algorithm with better scalability.

Findings

Unified multiple LKL approaches under one system

Developed a new algorithm with improved performance

Enhanced existing algorithms for scalability

Abstract

Multiple Kernel Learning, or MKL, extends (kernelized) SVM by attempting to learn not only a classifier/regressor but also the best kernel for the training task, usually from a combination of existing kernel functions. Most MKL methods seek the combined kernel that performs best over every training example, sacrificing performance in some areas to seek a global optimum. Localized kernel learning (LKL) overcomes this limitation by allowing the training algorithm to match a component kernel to the examples that can exploit it best. Several approaches to the localized kernel learning problem have been explored in the last several years. We unify many of these approaches under one simple system and design a new algorithm with improved performance. We also develop enhanced versions of existing algorithms, with an eye on scalability and performance.