What is CuBANz
CuBANz is a photometric redshift estimator code for high redshift galaxies which is created by Saumyadip Samui and Shanoli Samui Pal and available freely. It uses the back propagation neural network along with clustering of the training set, which makes it more efficient than other neural network codes. In CuBANz, the training set is divided into several self learning clusters with galaxies having similar photometric properties and spectroscopic redshifts within a given span. The clustering algorithm uses the colour information (i.e. u-g, g-r etc.) rather than the apparent magnitudes at various photometric bands as the photometric redshift is more sensitive to the flux differences between different bands rather than the actual values. Separate neural networks are trained for each clusters using all possible colors, magnitudes and uncertainties in the measurements.
For a galaxy with unknown redshift, CuBANz identify the closest possible clusters having similar photometric properties and use those clusters to get the photometric redshifts using the particular neural networks that were trained using those cluster members. For galaxies that do not match with any training cluster, the photometric redshifts are obtained from a separate network that uses entire training set. This clustering method enables us to
determine the redshifts more accurately.
Further, CuBANz provides a much better estimate of the uncertainty in the derived photometric redshift. It considers uncertainty in the photometric measurements as well as uncertainty in the neural network training.
The code is written in C language and can be run easily in any machine having standard C compiler. The source code can be downloaded from here.
Performance
CuBANz is tested with SDSS Stripe 82data. It produces the residual error <(zspec − zphot )2>1/2 = 0.03 with 3406 sources in the redshift range zspec < 0.7 having 5 bands ( u, g, r, i, z ) photometric measurements. The code performs very well even with 4 bands photometric measurements with residual error of 0.06. The banner plots show the performance of CuBANz for clustered sources (left) and sources that do not matches with training set (right) along with the predicted uncertainty.
See Samui & Pal, 2016, New Astronomy, 51 (2017), 169-177 for further details.