Projects, Talks, and Writing

Extracting Meaningful Features from Early-Science Radio Data

Early-science data from the Australian Square Kilometre Array Pathfinder (ASKAP) are coming fast. Wide-area radio projects such as the Evolutionary Map of the Universe (EMU) and the Polarisation Sky Survey of the Universe’s Magnetism (POSSUM) already have terabytes of ASKAP observations for use in early-science and survey planning. In this talk presented at C3DIS 2019 in Canberra and AI in Astronomy at ESO (Garching), I explain how I found a meaningful feature representation for prediction on early-science observations.

Radio luminosity functions with Radio Galaxy Zoo and machine learning

Legacy wide-area radio surveys have much science to offer, if we can get useful data out. I developed a machine learning approach to cross-identifying a large number of radio sources, and used this to make radio luminosity functions.

"Betwixt Odysseys: A Hellenic Soap Opera with a Godly Twist": A Critical Retrospective

I conduct a comical critical analysis on the talk immediately preceding mine at the 2019 Stromlo Student Seminars (itself by Adam Rains).

Big Radio Telescopes and Big Radio Galaxies

Next-generation radio telescopes will change radio astronomy. But what are next-generation radio telescopes? And what do we see in the radio sky? A talk presented at Smith's Alternative for the ANU Astronomy Society.

The Mysterious Radio Signal: An Unnatural Phenomenon?

I discuss the science of Sailor Scouts in this comedy talk for the Stromlo Student Seminars 2018.

Data challenges in citizen science

How does sourcing data from citizen science affect model validation? A talk presented at the Collaborative Conference on Computational and Data Intensive Science 2018 at the Melbourne Conference and Exhibition Centre.

Machine learning for radio cross-identification

Alger, M. J.; Banfield, J. K.; Ong, C. S.; Rudnick, L.; Wong, O. I.; Wolf, C.; Andernach, H.; Norris, R. P.; Shabala, S. S.

We present a machine learning approach for the task of determining the infrared host galaxies of radio emissions detected in wide-area radio surveys. Our method is trained on both expert cross-identifications and crowdsourced cross-identifications from Radio Galaxy Zoo and applied to observations from the Australia Telescope Large Area Survey.

I trained a classifier and now I don’t know what to do with it

Using classifiers for things that aren't classification. A talk presented at the 2017 Mount Stromlo Student Seminars at the Australian National University.

Acton

A Python package for active learning. Developed for the ARC Centre of Excellence for All-sky Astrophysics.

Crowdastro

Honours project at the Australian National University on machine learning applications in astronomy.

Question Time

An interactive heatmap of locations mentioned in the New South Wales parliament, with Hansard snippets and backlinks for particularly interesting places. GovHack 2016 project with Mitchell Busby.

Python: A Crash Course in Programming

Very (very) fast introductory scientific Python workshop I ran at Bruce Hall, Canberra.

Haskell: A Crash Course in Functional Programming

Very (very) fast introductory Haskell masterclass I ran at the National Computer Science School 2016, at the University of Sydney.

Inverse Reinforcement Learning

Python implementations of selected inverse reinforcement learning algorithms.

Coffeequate

Computer algebra system for JavaScript with limited equation solving and uncertainty propagation.

Circle Snake

Frustrating arcade-style browser game I made while learning JavaScript.