Evolving Neural Networks for the Classification of Galaxies

The FIRST survey (Faint Images of the Radio Sky at Twenty-cm) is scheduled to cover 10,000 square degrees of the northern and southern galactic caps. Until recently, astronomers classified radio-emitting galaxies through a visual inspection of FIRST images. Besides being subjective, prone to error and tedious, this manual approach is becoming infeasible: upon completion, FIRST will include almost a million galaxies. This paper describes the application of six methods of evolving neural networks (NNs) with genetic algorithms (GAs) to identify bent-double galaxies. The objective is to demonstrate that GAs can successfully address some common problems in the application of NNs to classification problems, such as training the networks, choosing appropriate network topologies, and selecting relevant features. The results indicate that most of the methods we tried performed equally well on our data, but using a GA to select features produced the best results