Artificial neural network to search for metal-poor galaxies^⋆

¹ North China Institute of Aerospace Engineering, Langfang, 065000 Hebei, PR China
e-mail: fshi@bao.ac.cn
² Center of Astrophysics, University of Science and Technology of China, Hefei, 230026, PR China
e-mail: xkong@ustc.edu.cn
³ Key Laboratory for Research in Galaxies and Cosmology, USTC, CAS, Hefei, 230026, PR China

Received: 5 September 2013
Accepted: 5 December 2013

Abstract

Aims. To find a fast and reliable method for selecting metal-poor galaxies (MPGs), especially in large surveys and huge databases, an artificial neural network (ANN) method is applied to a sample of star-forming galaxies from the Sloan Digital Sky Survey (SDSS) data release 9 (DR9) provided by the Max Planck Institute and the Johns Hopkins University (MPA/JHU).

Methods. A two-step approach is adopted: (i) The ANN network must be trained with a subset of objects that are known to be either MPGs or metal rich galaxies (MRGs), treating the strong emission line flux measurements as input feature vectors in n-dimensional space, where n is the number of strong emission line flux ratios. (ii) After the network is trained on a sample of star-forming galaxies, the remaining galaxies are classified in the automatic test analysis as either MPGs or MRGs. We consider several random divisions of the data into training and testing sets; for instance, for our sample, a total of 70 percent of the data are involved in training the algorithm, 15 percent are involved in validating the algorithm, and the remaining 15 percent are used for blind testing the resulting classifier.

Results. For target selection, we have achieved an acquisition rate for MPGs of 96 percent and 92 percent for an MPGs threshold of 12 +  log (O/H) = 8.00 and 12 +  log (O/H) = 8.39, respectively. Running the code takes minutes in most cases under the Matlab 2013a software environment. The ANN method can easily be extended to any MPGs target selection task when the physical property of the target can be expressed as a quantitative variable.