Centre de documentation
scientifique

Feature extractions for small sample size classification problem / B.C. Kuo in IEEE Transactions on geoscience and remote sensing, vol 45 n° 3 (March 2007)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 45 n° 3 (March 2007) . - pp 756 - 764 Titre : Feature extractions for small sample size classification problem Type de document : Article/Communication Auteurs : B.C. Kuo, Auteur ; K.Y. Chang, Auteur Année de publication : 2007 Article en page(s) : pp 756 - 764 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] algorithme génétique [Termes IGN] classification dirigée [Termes IGN] décomposition du pixel [Termes IGN] détection de contours [Termes IGN] reconnaissance de formes [Termes IGN] valeur propre Résumé : (Auteur) Much research has shown that the definitions of within-class and between-class scatter matrices and regularization technique are the key components to design a feature extraction for small sample size problems. In this paper, we illustrate the importance of another key component, eigenvalue decomposition method, and a new regularization technique was proposed. In the hyperspectral image experiment, the effects of these three components of feature extraction are explored on ill-posed and poorly posed conditions. The experimental results show that different regularization methods need to cooperate with different eigenvalue decomposition methods to reach the best performance, the proposed regularization method, regularized feature extraction (RFE) outperform others, and the best feature extraction for a small sample size classification problem is RFE with nonparametric weighted scatter matrices. Numéro de notice : A2007-088 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2006.885074 En ligne : https://doi.org/10.1109/TGRS.2006.885074 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=28453 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
065-07031	RAB	Revue	Centre de documentation	En réserve L003	Disponible

Nonparametric weighted feature extraction for classification / D.A. Landgrebe in IEEE Transactions on geoscience and remote sensing, vol 42 n° 5 (May 2004)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 42 n° 5 (May 2004) . - pp 1096 - 1105 Titre : Nonparametric weighted feature extraction for classification Type de document : Article/Communication Auteurs : D.A. Landgrebe, Auteur ; B.C. Kuo, Auteur Année de publication : 2004 Article en page(s) : pp 1096 - 1105 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image [Termes IGN] analyse discriminante [Termes IGN] classificateur non paramétrique [Termes IGN] extraction automatique [Termes IGN] extraction de traits caractéristiques [Termes IGN] matrice [Termes IGN] précision de la classification [Termes IGN] reconnaissance de formes [Termes IGN] réduction géométrique Résumé : (Auteur) In this paper, a new nonparametric feature extraction method is proposed for high-dimensional multiclass pattern recognition problems. It is based on a nonparametric extension of scatter matrices. There are at least two advantages to using the proposed nonparametric scatter matrices. First, they are generally of full rank. This provides the ability to specify the number of extracted features desired and to reduce the effect of the singularity problem. This is in contrast to parametric discriminant analysis, which usually only can extract L - 1 (number of classes minus one) features. In a real situation, this may not be enough. Second, the nonparametric nature of scatter matrices reduces the effects of outliers and works well even for nonnormal datasets. The new method provides greater weight to samples near the expected decision boundary. This tends to provide for increased classification accuracy. Numéro de notice : A2004-196 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2004.825578 En ligne : https://doi.org/10.1109/TGRS.2004.825578 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=26723 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
065-04051	RAB	Revue	Centre de documentation	En réserve L003	Disponible

A robust classification procedure based on mixture classifiers and nonparametric weighted feature extraction / B.C. Kuo in IEEE Transactions on geoscience and remote sensing, vol 40 n° 11 (November 2002)  

Public [article]  inIEEE Transactions on geoscience and remote sensing > vol 40 n° 11 (November 2002) . - pp 2486 - 2494 Titre : A robust classification procedure based on mixture classifiers and nonparametric weighted feature extraction Type de document : Article/Communication Auteurs : B.C. Kuo, Auteur ; D.A. Landgrebe, Auteur Année de publication : 2002 Article en page(s) : pp 2486 - 2494 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Photogrammétrie numérique [Termes IGN] classificateur non paramétrique [Termes IGN] classification [Termes IGN] extraction automatique [Termes IGN] extraction de traits caractéristiques [Termes IGN] image hyperspectrale [Termes IGN] méthode robuste Résumé : (Auteur) There are many factors to consider in carrying out a hyperspectral data classification. Perhaps chief among them are class training sample size, dimensionality, and distribution separability. The intent of this study is to design a classification procedure that is robust and maximally effective, but which provides the analyst with significant assists, thus simplifying the analyst's task. The result is a quadratic mixture classifier based on Mixed-LOOC2 regulized discriminant analysis and nonparametric weighted feature extraction. This procedure has the advantage of providing improved classification accuracy compared to typical previous methods but requires minimal need to consider the factor mentioned above. Experimental results demonstrating these properties are presented. Numéro de notice : A2002-359 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2002.805088 En ligne : https://doi.org/10.1109/TGRS.2002.805088 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=22270 [article]

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
065-02111	RAB	Revue	Centre de documentation	En réserve L003	Disponible