Feature Selection in Unsupervised Learning via Evolutionary Search

Feature Selection in Unsupervised Learning via Evolutionary Search (2000)

Cached

Download Links

[dollar.biz.uiowa.edu]

Other Repositories/Bibliography

DBLP

by Yongseog Kim , W. Nick Street , Filippo Menczer

Venue:	In Proceedings of the Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
Citations:	48 - 3 self

Versions

Version History

Metadata Version 3

User correction supplied by SystemCorrections

Datum	Value	Source
TITLE	Feature Selection in Unsupervised Learning via Evolutionary Search	user correction - Legacy Corrections
AUTHOR NAME	Yongseog Kim	user correction - Legacy Corrections
AUTHOR AFFIL	Management Sciences Department; University of Iowa	user correction - Legacy Corrections
AUTHOR ADDR	Iowa City, IA 52242 USA	user correction - Legacy Corrections
AUTHOR NAME	W. Nick Street	user correction - Legacy Corrections
AUTHOR AFFIL	Business Building, University of Iowa, Iowa City	user correction - Legacy Corrections
AUTHOR ADDR	IA 52242. Phone (319) 335-1016. Email	user correction - Legacy Corrections
AUTHOR NAME	Filippo Menczer	user correction - Legacy Corrections
ABSTRACT	Feature subset selection is an important problem in knowl- edge discovery, not only for the insight gained from deter- mining relevant modeling variables but also for the improved understandability, scalability, and possibly, accuracy of the resulting models. In this paper we consider the problem of feature selection for unsupervised learning. A number of heuristic criteria can be used to estimate the quality of clusters built from a given featuresubset. Rather than combining such criteria, we use ELSA, an evolutionary lo- cal selection algorithm that maintains a diverse population of solutions that approximate the Pareto front in a multi- dimensional objectiv espace. Each evolved solution repre- sents a feature subset and a number of clusters; a standard K-means algorithm is applied to form the given n umber of clusters based on the selected features. Preliminary results on both real and synthetic data show promise in finding Pareto-optimal solutions through which we can identify the significant features and the correct number of clusters.	user correction - Legacy Corrections
YEAR	2000	INFERENCE
VENUE	In Proceedings of the Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	365--369	INFERENCE
CITATIONS	31 found	ParsCit 1.0