Official source
Source domain: sarao.ac.za
Collected on 17 June 2026
Always confirm details on the official source before acting.
Read out of this document by Govermate's AI, in its own words. Each one quotes the line it came from — confirm on the official source before acting.
2027 MSc Project Proposal: Interpretable Machine Learning for Anomalous Radio Galaxy Morphologies
A 2027 MSc project is proposed to apply computer vision foundation models (such as DINOv3) to radio data and use interpretability tools (such as saliency maps/Grad-CAM) to understand what features of radio galaxies the model focuses on, using the MeerKAT Galaxy Cluster Legacy Survey (MGCLS) dataset and anomaly detection results from Astronomaly: Protege (Lochner & Rudnick 2025).
2027 MSc Project Proposal Cover page 1. Project Title Interpretable Machine Learning for Anomalous Radio Galaxy Morphologies
Places
Western Cape
Reference
PROPOSAL COVER PAGE 1
Opening Date
1 April 2026