标题：An Automatic Method for Selecting Special and Rare Celestial Objects in Massive Spectra
作者：Wang, Wenyu; Jiang, Bin
作者机构：[Wang, Wenyu; Jiang, Bin] Shandong Univ, Sch Mech Elect & Informat Engn, Weihai 264209, Peoples R China.
会议名称：Chinese Intelligent Automation Conference (CIAC)
来源：PROCEEDINGS OF THE 2015 CHINESE INTELLIGENT AUTOMATION CONFERENCE: INTELLIGENT INFORMATION PROCESSING
关键词：Data mining; Massive spectra; PCA; SVM
摘要：We proposed an automatic method for searching special and rare celestial objects in the massive spectra of the Sloan digital sky survey (SDSS). The data mining technique is employed and the massive SDSS spectra are identified quickly and efficiently. The high-dimensional spectra are mapped to feature space constructed by the principal component analysis (PCA), and dimensionality reduction is carried out accordingly. Massive SDSS spectra are classified by a well-trained support vector machine (SVM) and most of the noncandidates are excluded. Parameter optimization is also studied to guarantee the accuracy of PCA and SVM. Experiments show that this novel method can find rare celestial objects in an effective and efficient manner. We report the identification of six new white dwarf-main sequence (WDMS).