About me
I am a statistician working at the intersection of statistical methodology, machine learning and applied research. I currently work in industry on real-world health data, alongside my academic research as a Visiting Researcher at the Department of Computer Science, University of Oxford, and an Associate Researcher at Waseda University. Previously, I was a postdoctoral researcher in Professor Seth Flaxman’s group at Oxford. I completed my PhD in Statistics at the London School of Economics and Political Science (LSE), supervised by Professor Wicher Bergsma.
My research interests include Bayesian modelling, Gaussian processes, spatio-temporal analysis, small area estimation, survey statistics and interpretable statistical learning. I am particularly interested in applications to food security, public health and humanitarian research, and collaborate closely with the United Nations World Food Programme (WFP). I am also an active member of the Machine Learning and Global Health Network.
Projects
My methodological research focuses on developing flexible, interpretable and efficient statistical learning methods. My PhD research developed additive Gaussian process models, particularly for modelling and selecting interaction effects and for scalable inference with large or incomplete datasets. See this paper and these slides. More recently, I have been working on active learning and Bayesian optimisation for adaptive survey design, particularly to improve data collection when response rates and survey costs vary across locations and populations.
Much of my current research focuses on food security and nutrition, in collaboration with UN World Food Programme.
- Small area estimation of food insecurity: Combining mobile-phone and face-to-face surveys using Bayesian multilevel regression and post-stratification (MRP) to produce timely, population-representative estimates at fine geographical scales. See this paper.
- Mapping micronutrient inadequacy: Developing Bayesian small area estimation methods for estimating inadequate micronutrient intake at subnational levels, with applications in Rwanda, Senegal and Nigeria. See this paper.
- Adaptive learning for survey design (ALSD): Developing adaptive sampling strategies that use information from previous surveys to improve the representativeness and efficiency of future data collection.
- Spatio-temporal modelling of food insecurity: Developing scalable Gaussian process models for estimating and forecasting food insecurity in areas with limited or missing survey data. This is joint work with Dr. Francesca Panero and collaborators. See this paper.
Other things
Having suffered from eating disorder for a large part of my high school and undergraduate study and chronic fatigue syndrome during my PhD, I’d like to be understanding of everyone who is affected by health issues.
- I support BEAT, a UK based charity organisation, whenever I can.
- Any prospetive students with health issues considering doing a PhD, or others, are more than welcomed to reach out!