A new landmarker generation algorithm based on correlativity

Daren Ler*, Irena Koprinska, Sanjay Chawla

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Citations (Scopus)

Abstract

Landmarking is a recent and promising metalearning strategy, which defines meta-features that are themselves efficient learning algorithms. However, the choice of landmarkers is made in an ad hoc manner. In this paper, we propose a new perspective and set of criteria for landmarkers. With these, we introduce a landmarker generation algorithm, which creates a set of landmarkers that each utilise subsets of the algorithms being landmarked. The experiments show that the landmarkers formed, when used with linear regression, are able to estimate accuracy well, even when utilising a small fraction of the given algorithms.

Original languageEnglish
Title of host publicationProceedings of the 2004 International Conference on Machine Learning and Applications, ICMLA '04
EditorsM. Kantardzic, O. Nasraoui, M. Milanova
Pages178-185
Number of pages8
Publication statusPublished - 2004
Externally publishedYes
Event2004 International Conference on Machine Learning and Applications, ICMLA '04 - Louisville, KY, United States
Duration: 16 Dec 200418 Dec 2004

Publication series

NameProceedings of the 2004 International Conference on Machine Learning and Applications, ICMLA '04

Conference

Conference2004 International Conference on Machine Learning and Applications, ICMLA '04
Country/TerritoryUnited States
CityLouisville, KY
Period16/12/0418/12/04

Fingerprint

Dive into the research topics of 'A new landmarker generation algorithm based on correlativity'. Together they form a unique fingerprint.

Cite this