I'm a PhD student in the RainML lab at Oxford, supervised by Tom Rainforth and Adam Foster. My research has mainly focused on intelligent data acquisition. In machine learning we have lots of powerful methods for using data once we have it, yet we lack comparably capable methods for deciding what data to acquire in the first place. Fixing this shortfall could have huge impact: good data is crucial in model training and evaluation but also in more informally shaping decisions across society. Bayesian principles tells us how to find good data, and much of my work has been about translating those principles to state-of-the-art practical performance in modern contexts. Insights from my methods work have also informed some recent conceptual work rethinking a popular view on uncertainty.
