Automated Detection of Classical Novae with Neural Networks

The POINT-AGAPE collaboration surveyed M31 with the primary goal of optical detection of microlensing events, yet its data catalogue is also a prime source of lightcurves of variable and transient objects, including classical novae (CNe). A reliable means of identification, combined with a thorough survey of the variable objects in M31, provides an excellent opportunity to locate and study an entire galactic population of CNe. This paper presents a set of 440 neural networks, working in 44 committees, designed specifically to identify fast CNe. The networks are developed using training sets consisting of simulated novae and POINT-AGAPE lightcurves, in a novel variation on K-fold cross-validation. They use the binned, normalised power spectra of the lightcurves as input units. The networks successfully identify 9 of the 13 previously identified M31 CNe within their optimal working range (and 11 out of 13 if the network error bars are taken into account). They provide a catalogue of 19 new candidate fast CNe, of which 4 are strongly favoured.