Computer Science – Performance
Scientific paper
Jan 2007
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2007acasn..48....1z&link_type=abstract
Acta Astronomica Sinica, vol. 48, no. 1, p. 1-10
Computer Science
Performance
1
Galaxies: Active, Galaxies: Nuclei, Line: Profiles, Methods: Data Analysis
Scientific paper
Towards a set of low-redshift spectra of active galactic nuclei taken from the spectroscopic data of the Sloan Digital Sky Survey, a simple automated K-Nearest Neighbour method is developed to classify active galactic nuclei (AGN) into two types: broad-line AGNs and narrow-line AGNs. In this method, the wave bands of spectra in the rest frame are intercepted respectively corresponding to emission lines Hα, [OIII], Hα and [NII] because of the differences between the spectra of broad-line and narrow-line AGNs, and these features are used separately or together in the classifications. It is proved in the experiments that the optimal performance appears when just using the wave band of Hα and [NII] to carry out the classification, and the speed is 32.89s for single classification while the numbers of training and testing data are respectively 1000 and 3313. Moreover the Hα emission line appears the uppermost difference between the two types of AGNs. This research shows that under conditions of making full use of typical spectral features, automated classification method is feasible to classify the spectra of active galactic nuclei, and also provides a fast and straightforward alternative to the classifications using the FWHM (Full Width at Half Maximum Height) values of emission lines or line strength ratio diagnostic diagrams for large amount of spectra from the large-scale spectral surveys.
Luo A.-Li
Wu Congjun
Wu Feng-Cheng
Zhao Mei-Fang
Zhao Yong-Heng
No associations
LandOfFree
Automated spectral classification of broad-line and narrow-line active galactic nuclei based on K-nearest neighbour does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Automated spectral classification of broad-line and narrow-line active galactic nuclei based on K-nearest neighbour, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automated spectral classification of broad-line and narrow-line active galactic nuclei based on K-nearest neighbour will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1362859