Geodesic
On selecting the best features in a noisy environment
Kybernetika, Tome 34 (1998) no. 4, p. [411]

Voir la notice de l'article provenant de la source Czech Digital Mathematics Library

This paper introduces a novel method for selecting a feature subset yielding an optimal trade-off between class separability and feature space dimensionality. We assume the following feature properties: (a) the features are ordered into a sequence, (b) robustness of the features decreases with an increasing order and (c) higher-order features supply more detailed information about the objects. We present a general algorithm how to find under those assumptions the optimal feature subset. Its performance is demonstrated experimentally in the space of moment-based descriptors of 1-D signals, which are invariant to linear filtering.
Classification : 62H30, 62H99, 62M20, 65C60, 68T10
Keywords: Mahalanobis distance; 1-D signals 
@article{KYB_1998__34_4_a9,
     author = {Flusser, Jan and Suk, Tom\'a\v{s}},
     title = {On selecting the best features in a noisy environment},
     journal = {Kybernetika},
     pages = {[411]},
     publisher = {mathdoc},
     volume = {34},
     number = {4},
     year = {1998},
     zbl = {1274.62433},
     language = {en},
     url = {http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a9/}
}
TY  - JOUR
AU  - Flusser, Jan
AU  - Suk, Tomáš
TI  - On selecting the best features in a noisy environment
JO  - Kybernetika
PY  - 1998
SP  - [411]
VL  - 34
IS  - 4
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a9/
LA  - en
ID  - KYB_1998__34_4_a9
ER  - 
%0 Journal Article
%A Flusser, Jan
%A Suk, Tomáš
%T On selecting the best features in a noisy environment
%J Kybernetika
%D 1998
%P [411]
%V 34
%N 4
%I mathdoc
%U http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a9/
%G en
%F KYB_1998__34_4_a9
Flusser, Jan; Suk, Tomáš. On selecting the best features in a noisy environment. Kybernetika, Tome 34 (1998) no. 4, p. [411]. http://geodesic.mathdoc.fr/item/KYB_1998__34_4_a9/