Research

My work focuses on optimal policy learning (i.e. optimal treatment regimens), treatment effect heterogeneity and counterfactual prediction. I am particularly interested in orthogonalised estimators (exploiting influence-functions).

I am also co-lead in a collaborative research project “Developing statistical machine learning methods for Clinical Trials” based at the Alan Turing Institute. We are building machine learning models to predict trial site risk of under-performing (in the sense of low quality data, high rate of missing data or higher than expected adverse events, for example).

I am interested in studying fairness and explainability of prediction models and heterogeneous treatments effects (through the lens of causal inference).

PhD supervision

I am happy to supervise new PhD students starting in September 2024.

Current students:

Matt Pryce (2021 - present): MRC London Intercollegiate Doctoral scholarship studying double-robust methods with machine learning for survival analysis in high dimensional settings.

Rebecca Xu (2021- present) : ESRC Doctoral Scholarship studying extensions to target trial emulation for group policy evaluation, in collaboration with the Health Foundation.

Former PhD students:

Dr Oliver Hines (awarded 2023): ``Assumption-Lean Inference for Causal and Statistical Questions in the Era of Machine

Learning’’ Funded by an MRC London Intercollegiate Doctoral scholarship (jointly supervised with Prof Stijn Vansteelandt)

Dr Schadrac Agbla (awarded 2019): Instrumental Variable methods for adjusting for nonadherence in cluster randomised trials (joint with Prof Bianca DeStavola).

Dr Anower Hossain (awarded 2017): Missing data methodology for cluster randomised trials (joint with Dr Jonathan Bartlett).

Other

I am a member of the steerting committee for the European Causal Inference Meeting (EuroCIM).