Journal article
Evolutionary model selection in unsupervised learning
Intelligent data analysis, Vol.6(6), pp.531-556
2002
DOI: 10.3233/IDA-2002-6605
Abstract
Feature subset selection is important not only for the insight gained from determining relevant modeling variables but also for the improved understandability, scalability, and possibly, accuracy of the resulting models. Feature selection has traditionally been studied in supervised learning situations, with some estimate of accuracy used to evaluate candidate subsets. However, we often cannot apply supervised learning for lack of a training signal. For these cases, we propose a new feature selection approach based on clustering. A number of heuristic criteria can be used to estimate the quality of clusters built from a given feature subset. Rather than combining such criteria, we use ELSA, an evolutionary local selection algorithm that maintains a diverse population of solutions that approximate the Pareto front in a multi-dimensional objective space. Each evolved solution represents a feature subset and a number of clusters; two representative clustering algorithms, K-means and EM, are applied to form the given number of clusters based on the selected features. Experimental results on both real and synthetic data show that the method can consistently find approximate Pareto-optimal solutions through which we can identify the significant features and an appropriate number of clusters. This results in models with better and clearer semantic relevance.
Details
- Title: Subtitle
- Evolutionary model selection in unsupervised learning
- Creators
- YongSeog Kim - Utah State UniversityW. Nick Street - Utah State UniversityFilippo Menczer - Utah State University
- Resource Type
- Journal article
- Publication Details
- Intelligent data analysis, Vol.6(6), pp.531-556
- DOI
- 10.3233/IDA-2002-6605
- ISSN
- 1088-467X
- eISSN
- 1571-4128
- Language
- English
- Date published
- 2002
- Academic Unit
- Bus Admin College; Nursing; Business Analytics; Computer Science
- Record Identifier
- 9984380387702771
Metrics
7 Record Views