A simple and robust method for automated photometric classification of supernovae using neural networks

A method is presented for automated photometric classificat ion of supernovae (SNe) as TypeIa or non-Ia. A two-step approach is adopted in which: (i) the SN lightcurve flux measurements in each observing filter are fitted separately to an anal ytical parameterised function that is sufficiently flexible to accommodate vitrually all types o f SNe; and (ii) the fitted function parameters and their associated uncertainties, along with the number of flux measurements, the maximum-likelihood value of the fit and Bayesian evidenc e for the model, are used as the input feature vector to a classification neural network ( NN) that outputs the probability that the SN under consideration is of Type-Ia. The method is trained and tested using data released following the SuperNova Photometric Classificati on Challenge (SNPCC), consisting of lightcurves for 21,319 SNe in total. We consider several r andom divisions of the data into training and testing sets: for instance, for our sample D1 (D4), a total of 10 (40) per cent of the data are involved in training the algorithm and the remai nder used for blind testing of the resulting classifier; we make no selection cuts. Assigning a canonical threshold probability of pth = 0.5 on the network output to class a SN as Type-Ia, for the sample D1 (D4) we obtain a completeness of 0.78 (0.82), purity of 0.77 (0.82), an d SNPCC figure-of-merit of 0.41 (0.50). Including the SN host-galaxy redshift and its uncer tainty as additional inputs to the classification network results in a modest 5‐10 per cent incr ease in these values. We find that the quality of the classification does not vary significantly with SN redshift. Moreover, our probabilistic classification method allows one to calculat e the expected completeness, purity and figure-of-merit (or other measures of classification qua lity) as a function of the threshold probability pth, without knowing the true classes of the SNe in the testing sample, as is the case in the classification of real SNe data. The method may thus be improved further by optimising pth and can easily be extended to divide non-Ia SNe into their different classes.