Research interests: Computational statistics, renewable energy, machine/deep learning.
My PhD focuses on using statistics to make the best decisions in the energy transition, particularly in the electricity sector. Such decisions (e.g. whether to build a wind farm, battery or new transmission line) are typically difficult due to the complexity of electricity grids and considerable uncertainty regarding future government policy, grid developments, electricity demand/price patterns and weather events.
For this reason, electricity strategy is typically informed by large amounts of data. For example, to determine the best location of a new wind farm, historic wind speeds and electricity prices may be employed, along with a model of the grid. This leads to a natural statistics problem: how do we use the available data to make the best decisions in the energy transition?
See Google Scholar page for up-to-date publication list.
AP Hilbers, DJ Brayshaw, A Gandy (2020). Efficient quantification of the impact of demand and weather uncertainty in power system models. To appear in IEEE Transactions on Power Systems. [preprint] [code]
AP Hilbers, DJ Brayshaw, A Gandy (2020). Importance subsampling for power system planning under multi-year demand and weather uncertainty1. In proceedings of the 16th International Conference on Probabilistic Methods Applied to Power Systems (PMAPS 2020). [paper] [open access version] [code]
1Runner-up for Roy Billinton Award for best student paper
2See 2020 importance subsampling paper for generalised version
I helped produce a Coursera course in Tensorflow 2, which came out in 2020. As of October 2020, over 10,000 people have taken the course and it has a rating of 4.9/5. It’s free to take for anyone! Check it out here.
|2019||Machine Learning (university level summer school)|
|2019||Machine Learning (master level data science for healthcare course)|
|2019||Mathematical Methods (first year undergraduate mathematics course)|
|2018||Data & Uncertainty (master’s level mathematics course)|
All courses at Imperial College London.