Writing
I have written about machine learning, statistics and scientific computing since 2019.
researchdatapod.com
The Research Scientist Pod is where I write in public. It began as a place to discuss recent research in machine learning, then branched out into tutorials, visualisations and statistics calculators. The readers are technical people who want to understand a method well enough to use it themselves, rather than have an AI intermediary use it for them.
Selected pieces
- An analysis of the scaling behaviour of GPT-3, published in August 2020, written while the question of what scale alone could buy was still open.
- The history of reinforcement learning, tracing the field from early animal learning experiments to deep reinforcement learning and its use in large language models.
- A guide to the Bellman-Ford algorithm, covering the theory, the mathematics and working implementations in Python, Java, C++ and JavaScript.
Peer-reviewed publications
I have four first-author publications with the IEEE, on log message classification using language modelling and word embeddings, and on anomaly detection including fault localisation and periodicity analysis. In 2023 I co-chaired the AnNet workshop at IEEE/IFIP NOMS.