Hunting the Unknowns in Murriyang, an Anomaly Detection Approach
Dr Suk Yee Yong1
How to search for something if we don’t know what we are looking for? In the era of big data astronomy, astronomical data are already generating large volume of data in the terabytes and petabytes scales. This includes the Murriyang (or Parkes) telescope, which has been mapping the sky in search for radio signals. Although current traditional search method has been successful in discovering many radio sources, it also yield many false candidates and only applicable to specific known signals. Using an anomaly detection method coupled with simulation data, we aim to detect any signals that could be astronomical or from possible extraterrestrial intelligence. We present a comparison between our approach and the standard search method.
My current research involves finding the unknowns in astronomical dataset using machine learning. My research interests include development and application of machine learning and deep learning techniques to solve real world problems.