Back to Volume
Paper: Discrimination of Point Sources Using Radial Basis Function Network and Random Forest
Volume: 394, Astronomical Data Analysis Software and Systems (ADASS) XVII
Page: 543
Authors: Zhang, Y.; Zhao, Y.; Gao, D.
Abstract: Selecting nonrepeat photometric objects (41,359 stars and 43,010 quasars) from the Second Data Release of the sloan digital sky survey (SDSS), we apply Radial Basis Function (RBF) network and Random Forest to separate quasars from stars. RBF network implements a normalized Gaussian radial basis function network. It uses the k-means clustering algorithm to provide the basis functions and learns a linear regression (numeric class problems) on top of that. Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The experimental results show that the accuracy of RBF network and Random Forest adds up to 93.71% and 96.98%, respectively. Apparently these two methods are applicable and effective to classify quasars from stars. They may also be used to solve other classification problems faced in astronomy and Virtual Observatory.
Back to Volume