Stability of feature selection algorithms: a study on high-dimensional spaces

Kalousis, Alexandros; Prados, Julien; Hilario, Mélanie

doi:10.1007/s10115-006-0040-8

Cited by 535 publications

(379 citation statements)

References 12 publications

Supporting

Mentioning

372

Contrasting

Unclassified

Order By: Relevance

“…Subsequently, feature selection is performed on each of the k subsamples, and a measure of stability or robustness is calculated. Here, following [5], we take a similarity based approach where feature stability is measured by comparing the outputs of the feature selectors on the k subsamples. The more similar all outputs are, the higher the stability measure will be.…”

Section: Estimating Stability With Instance Perturbationmentioning

confidence: 99%

“…Recent work in this area mainly focuses on the stability indices to be used for feature selection, introducing measures based on Hamming distance [4], correlation coefficients [5], consistency [6] and information theory [7]. Kalousis and coworkers also present an extensive comparative evaluation of feature selection stability over a number of high-dimensional datasets [5]. However, most of this work only focuses on the stability of single feature selection techniques, an exception being the work of [4] which describes an example combining multiple feature selection runs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Robust Feature Selection Using Ensemble Feature Selection Techniques

Saeys

Abeel

Peer

2008

Lecture Notes in Computer Science

537

393

View full text Add to dashboard Cite

Abstract. Robustness or stability of feature selection techniques is a topic of recent interest, and is an important issue when selected feature subsets are subsequently analysed by domain experts to gain more insight into the problem modelled. In this work, we investigate the use of ensemble feature selection techniques, where multiple feature selection methods are combined to yield more robust results. We show that these techniques show great promise for high-dimensional domains with small sample sizes, and provide more robust feature subsets than a single feature selection technique. In addition, we also investigate the effect of ensemble feature selection techniques on classification performance, giving rise to a new model selection strategy.

show abstract

Section: Estimating Stability With Instance Perturbationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Robust Feature Selection Using Ensemble Feature Selection Techniques

Saeys

Abeel

Peer

2008

Lecture Notes in Computer Science

537

393

View full text Add to dashboard Cite

show abstract

“…Table 6 lists one stability measure MW1 [72]. The Pearson's correlation coefficient ranges from -1 to 1.…”

Section: Weighting-score Vectormentioning

confidence: 99%

Stable feature selection for biomarker discovery

2010

Computational Biology and Chemistry

276

186

View full text Add to dashboard Cite

Feature selection techniques have been used as the workhorse in biomarker discovery applications for a long time. Surprisingly, the stability of feature selection with respect to sampling variations has long been under-considered. It is only until recently that this issue has received more and more attention. In this article, we review existing stable feature selection methods for biomarker discovery using a generic hierarchal framework. We have two objectives: (1) providing an overview on this new yet fast growing topic for a convenient reference; (2) categorizing existing methods under an expandable framework for future research and development.

show abstract

“…Several similarity measures have been proposed in the literature. Some of the older works propose using the Jaccard index sim J [6] (also referred as the Tanimoto distance) or the relative Hamming distance to define a similarity measure sim H [4]. Let us assume that we have n features in total.…”

Section: Existing Measuresmentioning

confidence: 99%

Measuring the Stability of Feature Selection with Applications to Ensemble Methods

Nogueira

Brown

2015

Multiple Classifier Systems

View full text Add to dashboard Cite

Abstract. Ensemble methods are often used to decide on a good selection of features for later processing by a classifier. Examples of this are in the determination of Random Forest variable importance proposed by Breiman, and in the concept of feature selection ensembles, where the outputs of multiple feature selectors are combined to yield more robust results. All of these methods rely critically on the concept of feature selection stability -similar but distinct to the concept of diversity in classifier ensembles. We conduct a systematic study of the literature, identifying desirable/undesirable properties, and identify a weakness in existing measures. A simple correction is proposed, and empirical studies are conducted to illustrate its utility.

show abstract

Stability of feature selection algorithms: a study on high-dimensional spaces

Abstract: With the proliferation of extremely high-dimensional

Cited by 535 publications

References 12 publications

Robust Feature Selection Using Ensemble Feature Selection Techniques

Robust Feature Selection Using Ensemble Feature Selection Techniques

Stable feature selection for biomarker discovery

Measuring the Stability of Feature Selection with Applications to Ensemble Methods

Contact Info

Product

Resources

About