Conference paper

ŠKODA Petr, VRÁBELOVÁ Pavla and ZENDULKA Jaroslav. Feature Extraction Using Wavelet Power Spectrum for Stellar Spectra Clustering. In: Proceedings of the 11th annual conference Znalosti 2012. Praha, 2012, pp. 31-40. ISBN 978-80-7378-220-7.
Publication language:english
Original title:Feature Extraction Using Wavelet Power Spectrum for Stellar Spectra Clustering
Pages:31-40
Proceedings:Proceedings of the 11th annual conference Znalosti 2012
Conference:ZNALOSTI 2012
Place:Praha, CZ
Year:2012
ISBN:978-80-7378-220-7
Files: 
+Type Name Title Size Last modified
iconllncs.pdf2,74 MB2012-05-29 12:21:45
^ Select all
With selected:
Abstract
This paper analyses the capabilities of using wavelet power spectrum for clustering of Be-type stars spectra. We propose a method using discrete wavelet transform for feature extraction and the wavelet power spectrum as a feature vector. We also propose a modification of this method and compare them. We analyse the methods in the clustering of artificial stellar spectra and compare them with a traditional method of wavelet-based feature extraction -- keeping $k$ largest coefficients. The results show that the correctness of clustering of our method is significantly better than in the case of a traditional method. We also compare the effect of using different type of wavelet and level of decomposition.

BibTeX:
@INPROCEEDINGS{
   author = {Petr {\v{S}}koda and Pavla Vr{\'{a}}belov{\'{a}} and
	Jaroslav Zendulka},
   title = {Feature Extraction Using Wavelet Power Spectrum for Stellar
	Spectra Clustering},
   pages = {31--40},
   booktitle = {Proceedings of the 11th annual conference Znalosti 2012},
   year = {2012},
   location = {Praha, CZ},
   ISBN = {978-80-7378-220-7},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=10061}
}

Your IPv4 address: 54.80.211.135
Switch to IPv6 connection

DNSSEC [dnssec]