This project is an attempt to create a machine-learning algorithm (binary classifier) that will be able to tell whether a polarized radio spectrum is a simple "Faraday thin" source or a more complex source with multiple components or "Faraday thick". The classifier is a convolutional neural network (faraday_cnn.py), which uses the faraday spectrum as the input feature vector. This is all currently work-in-progress. The work is in collaboration with the POSSUM survey with the Australian Square Kilometre Array Pathfinder (ASKAP). Bellow is an example of a simple Faraday thin source produced with create_spectrum.py
Here is a two component source, also created with create_spectrum.py, that has one component at 7 rad/m^2 and another at 40 rad/m^2. The frequency coverage mimics what will be available for POSSUM Early Science, though there could be an option to fill in the missing frequencies in some cases.
