Talks

• Scaling R with AWS Lambda

In this talk I gave an overview of a recent client project in which AWS Lambda was used to provide a scalable R-backend for a public-facing web application. This backend performed a number of different operations, including; evaluating a Bayesian Network model; rendering a parameterised PDF report via R Markdown, and creating data visualisations with {ggplot2}. The frontend and backend code for this application is available publicly on GitHub here. Similarly, the infrastructure as code to host the front and backend is available here. I gave this talk at the North East Data Science meetup in September 2022.


• Ants, Circle Pits, & Peter Jackson

I gave this talk at Newcastle University’s postgraduate forum seminar series. This seminar series is intended to be more relaxed than a more traditional academic seminar environment. Speakers are encouraged to lighten the mathematical load, and deliver a talk that is accessible to mathematicians from all disciplines. NB. The original slides were presented in reveal.js format. However, GitHub’s dependabot was continually unhappy with the accompanying package-lock.json file. In an attempt to permanently silence dependabot, I have converted these reveal.js slides into PDF format.


• Bayesian statistics in Astrophysics

I delivered this talk to Newcastle University’s applied mathematics department in October 2019. The intention of this talk was to introduce attendees to Bayesian statistics, highlight how it differs from frequentist statistics and provide an astronomically motivated case study. All the code associated with these slides can be found in this repository


• Hash functions and password cracking

Cryptographic hash functions have been described as the workhorses of modern cryptography, having uses in message authentication, digital signatures and password verification. But how do they work and how safe really are they? In this talk I gave a live demonstration of password cracking using the software hashcat and one of the university’s GPU equipped machines. This machine allowed us to make over 4 billion password attacks per second. I concluded the talk by discussing how to set and manage passwords in a secure and safe manner.


• Parameter inference for collective behaviour

In this talk, given in 2018 at Newcastle University, I discussed work I was completing as part of my PhD. In particular, I gave a brief introduction to the phenomenon of collective behaviour and the related mathematical models used to simulate these behaviours. I then showed how we can use the Stan programming language to learn about individual behaviours from field data.