Publication Details
Classification of Spectra of Emission-line Stars Using Feature Extraction Based on Wavelet Transform
VRÁBELOVÁ Pavla, BAŘINA David, ŠKODA Petr, VÁŽNÝ Jaroslav and ZENDULKA Jaroslav. Classification of Spectra of Emission-line Stars Using Feature Extraction Based on Wavelet Transform. In: Proceedings of 23rd Annual Astronomical Data Analysis Software and Systems (ADASS) Conference. San Francisco: Astronomical Society of the Pacific, 2013, pp. 177-180. ISBN 978-1-58381-854-1. Available from: http://www.aspbooks.org/publications/485/177.pdf
Czech title
Klasifikace spekter emisních čar hvězd použitím extrakce příznaků založené na vlnkové transformaci
Type
conference paper
Language
english
Authors
Vrábelová Pavla, Ing. (DIFS FIT BUT)
Bařina David, Ing., Ph.D. (DCGM FIT BUT)
Škoda Petr (ASU)
Vážný Jaroslav, Mgr. (MUNI)
Zendulka Jaroslav, doc. Ing., CSc. (DIFS FIT BUT)
Bařina David, Ing., Ph.D. (DCGM FIT BUT)
Škoda Petr (ASU)
Vážný Jaroslav, Mgr. (MUNI)
Zendulka Jaroslav, doc. Ing., CSc. (DIFS FIT BUT)
URL
Keywords
discrete wavelet transform, DWT, stationary wavelet transform, SWT, support vector machines, SVM
Abstract
The paper deals with classification of Be star spectra using discrete wavelet transform (DWT), stationary wavelet transform (SWT) and support vector machines (SVM).
Annotation
Our goal is to automatically identify Be and B[e] stars spectra in large archives and classify their types. Due to the length of spectra, the dimension reduction is required. As the Be stars show a number of different shapes of emission lines, it is not easy to construct simple criteria to identify the emission lines. We propose to perform the wavelet transform and apply statistical metrics on the wavelet coefficients. The results show that using proposed method we can reduce the number of attributes and the processing time to a small fraction, and moreover increase the accuracy.
Published
2013
Pages
177-180
Proceedings
Proceedings of 23rd Annual Astronomical Data Analysis Software and Systems (ADASS) Conference
Conference
Astronomical Data Analysis Software and Systems, Waikoloa, Hawaii, US
ISBN
978-1-58381-854-1
Publisher
Astronomical Society of the Pacific
Place
San Francisco, US
UT WoS
000376047600040
BibTeX
@INPROCEEDINGS{FITPUB10412, author = "Pavla Vr\'{a}belov\'{a} and David Ba\v{r}ina and Petr \v{S}koda and Jaroslav V\'{a}\v{z}n\'{y} and Jaroslav Zendulka", title = "Classification of Spectra of Emission-line Stars Using Feature Extraction Based on Wavelet Transform", pages = "177--180", booktitle = "Proceedings of 23rd Annual Astronomical Data Analysis Software and Systems (ADASS) Conference", year = 2013, location = "San Francisco, US", publisher = "Astronomical Society of the Pacific", ISBN = "978-1-58381-854-1", language = "english", url = "https://www.fit.vut.cz/research/publication/10412" }
Files