Classifying complex Faraday spectra with convolutional neural networks

Advances in radio spectropolarimetry offer the possibility to disentangle complex regions where relativistic and thermal plasmas mix in the interstellar and intergalactic media. Recent work has shown that apparently simple Faraday rotation measure spectra can be generated by complex sources. This is true even when the distribution of rotation measures in the complex source greatly exceeds the errors associated with a single component fit to the peak of the Faraday spectrum. We present a convolutional neural network that can differentiate between simple Faraday thin spectra and those that contain multiple (two) Faraday thin sources. We demonstrate that this network, trained for the upcoming Polarization Sky Survey of the Universe's Magnetism early science observations, can identify two component sources 99 per cent of the time, provided that the sources are separated in Faraday depth by >10 per cent of the full width at half-maximum of the Faraday point spread function, the polarized flux ratio of the sources is >0.1, and that the signal-to-noise ratio (S/N) of the primary component is >5. With this S/N cut-off, the false positive rate (simple sources misclassified as complex) is <0.3 per cent. Work is ongoing to include Faraday thick sources in the training and testing of the convolutional neural network.