Next-Generation Radio Surveys for Cosmic Magnetism Science

In this project the student will work on spectral data from the MeerKAT MIGHTEE survey, with particular emphasis on the polarisation components, to produce a classified object catalogue with quantitative estimates of bias. In collaboration with the MIGHTEE team, the student will develop the polarisation calibration and imaging processing for MeerKAT survey data to produce a classified object catalogue with quantitative estimates of bias using machine learning methodologies. A key focus within the later part of this project will be understanding the effects of selection bias in the training data for the machine learning classification and the impact of these biases on astrophysical interpretation and parameter estimation.