Organised by:
in collaboration with:

secretariat:

 

social media:


 



Accesso Utenti

Contact image
National Astronomical Observatories, CAS
Address:
China
Miscellaneous Information:

Abstract Reference: 30785
Identifier: P1.33
Presentation: Poster presentation
Key Theme: 1 Reduction and Analysis Algorithms for Large Databases and Vice-versa

Extra-Trees for Photometric Redshift Estimation of Quasars

Authors:
Zhang Yanxia, Yang Tu, Zhao Yongheng, Tian Haijun

Based on the Sloan Digital Sky Survey (SDSS) DR7 and DR12, UKIRT Infrared Deep Sky Survey (UKIDSS) and Wide-field Infrared Survey Explorer (WISE), we obtain different cross-matched samples and use a kind of tree-based method, extremely randomized trees (Extra-Trees) to estimate the photometric redshifts of quasars, moreover compare the performance of this method with k-nearest neighbor algorithm (KNN). Our experimental results show that the accuracy of predicting photometric redshifts is influenced by many factors, such as the sample quality, sample selection, feature selection and adopted algorithms. Optimal selection of samples and features contributes to the performance improvement of a regressor. Extra-Trees get better performance than KNN in the low dimensional space while KNN is superior to Extra-Trees in the high dimensional space.