1. Base Department, Beijing Institute of Clothing Technology,Beijing 100029, China 2. Department of Computer, Xi’an University of Science and Technology, Xi’an 710054, China 3. College of Information Science and Technology, Beijing Normal University,Beijing 100875, China
Abstract:The rapid development of astronomical observation has led to many large sky surveys such as SDSS (Sloan digital sky survey) and LAMOST (large sky area multi-object spectroscopic telescope). Since these surveys have produced very large numbers of spectra, automated spectral analysis becomes desirable and necessary. The present paper studies the spectral classification method based on Bayes decision theory, which divides spectra into three types: star, galaxy and quasar. Firstly, principal component analysis (PCA) is used in feature extraction, and spectra are projected into the 3D PCA feature space; secondly, the class conditional probability density functions are estimated using the non-parametric density estimation technique, Parzen window approach; finally, the minimum error Bayes decision rule is used for classification. In Parzen window approach, the kernel width affects the density estimation, and then affects the classification effect. Extensive experiments have been performed to analyze the relationship between the kernel widths and the correct classification rates. The authors found that the correct rate increases with the kernel width being close to some threshold, while it decreases with the kernel width being less than this threshold.
[1] Zhao Yong-heng. Modern Physics, 2007,19(5): 3. [2] Bai L, Guo P, Hu Z Y. China Journal of Astronomy and Astrophysics,2005, 5(2): 203. [3] Bailer-Jones C, Irwin M, Von Hippel T. Monthly Notices of Royal Astronomical Society, 1998, 298(2): 361. [4] Von Hippel T, Storrie-Lombardi L J, Storrie-Lombardi M C, et al. Monthly Notices of Royal Astronomical Society, 1994, 269(1): 97. [5] Connolly A J, Szalay A S, Bershady M A, et al. Astronomical Journal, 1995, 110(3): 1071. [6] Gaspar Galaz, Valerie de Lapparent. Astronomy and Astrophysics, 1998, 332(2): 459. [7] Zaritsky D, Zabludoff A I, Jeffrey A W. Astronomical Journal, 1995, 110(7): 1602. [8] LIU Rong, DUAN Fu-qing, LIU San-yang, et al(刘 蓉,段福庆,刘三阳,等). Acta Electronica Sinica(电子学报), 2005, 33(11): 2059. [9] Qin D M, Guo P, Hu Z Y, et al. China Journal of Astronomy and Astrophysics, 2003, 3(3): 277. [10] Zhang Y,Zhao Y. Astronomy and Astrophysics,2004, 422(3): 1113. [11] Zhang Y, Zhao Y. PASP, 2003, 115(8): 1006. [12] YANG Jin-fu, XU Xin, WU Fu-chao, et al(杨金福, 许 馨, 吴福朝,等). Spectroscopy and Spectral Analysis(光谱学与光谱分析) 2007, 27(3): 602. [13] BIAN Zhao-qi, ZHANG Xue-gong(边肇祺, 张学工). Pattern Recognition(模式识别). Beijing: Tsinghua University Press (北京: 清华大学出版社), 2000. [14] Glazebrook K, Offer A R, Deeley K. Astrophysical Journal, 1998, 492(1): 98. [15] Parzen E. Annals of Mathematical Statistics, 1962, 33(3): 1065. [16] Jones M C. Statistics and Probability Letters, 1990,9: 129. [17] Altman Naomi,Christian Leger. Journal of Statistical Planning and Inference, 1995, 46: 195. [18] Patrick D G, William R S. Biometrics, 2004, 55(3): 769.