Reviews, lecture notes and technical articles in physics, machine learning and mathematics. Each entry opens with a short abstract.
Research interests, with links to the related articles and projects on this site.
From fundamental questions to engineering practice: quantum fields in curved spacetime, black holes and matrix models of M-theory, particle phenomenology at colliders and in dark-matter searches, and analytical models for real engineering problems, such as pipeline corrosion.
Machine learning as a scientific instrument, in physics, medicine and other fields where data must serve as evidence, with emphasis on calibration, validation and a clear statement of each model's limitations.
Learning systems seen as complex systems: how loss landscapes shape what gradient-based optimisation can find, how neural networks store and retrieve memories, and the point at which a signal becomes learnable from noisy, high-dimensional data.
Remote sensing and geospatial machine learning: land-cover and urban-growth mapping, monitoring of crops, water and ecosystems, and assessment of map accuracy.