Homepage

Using data driven methods from statistics and machine learning, my work lies at the intersection between social science and computer science.

Recently, I have been working quite a bit with neural networks to study data privacy and institutions in governance. Some of my projects have had a particular focus on Latin America.

You can read my Political Science papers in the Journal of Politics or International Interactions, among others. I presented findings relevant to Computer Science at the International Conference on Machine Learning and the Theory and Practice of Differential Privacy Workshop Series.

I regularly collaborate with private and public institutions as a consultant.

News

March 2021 The programme for the 1st PolMeth Europe is online now.

February 2021 I will be teaching the masterclass on Deep Learning at the Q-Step Centre in Warwick again this year. The class is on Feb. 22nd–all online of course. To book, follow this link.

February 2021 ICLR 2021will have a workshop on the quality, privacy and bias of synthetic data. Submit your proposals until Feb 26th here. I will be on the Programme Committee and am looking forward to reading loads of SocSci related work.

February 2021 I am teaching the fully flipped ‘Intro to Data Science’ for the first time this year to Cardiff’s undergraduates. This way to the module home.

January 2021 Michelle Brown (NDI), J. Andrew Harris (NYU Abu Dhabi), Zach Warner (Cardiff) and myself have been developing CNNs to detect voting irregularities via image data (documents & satellites). We just received ESRC funding that helps us putting things into practice at the National Democratic Institute.

November 2020 Marcel Neunhoeffer (Mannheim) and I presented our paper Really Useful Synthetic Data — Promises and Challenges of Releasing Sensitive Information With Differentially Private Data Synthesizers at this year’s TPDP as part of CCS 2020 and also at the AI Policy Conference at RegHorizon.

August 2020 I was promoted to Senior Lecturer this year. Science is always a team effort, so a big ‘thank you’ to all collaborators.

July 2020 Marcel Neunhoeffer (Mannheim) and I will present our working paper Really Useful Synthetic Data — Promises and Challenges of Releasing Sensitive Information With Differentially Private Data Synthesizers at this year’s ICML in the EcoPaDL workshop. [paper]

July 2020 Our paper Hidden in Plain Sight? Detecting Electoral Fraud Using Statutory Results with Michelle Brown (NDI), J. Andrew Harris (NYU AD) and Zach Warner (Cardiff) is at this year’s XXXVII Polmeth. [paper]

June 2020 Marcel Neunhoeffer (Mannheim) and I presented our working paper on the utility of differentially private synthetic data at Microsoft. [presentation, paper]

June 2020 I spoke at CogX: “Using Artificial Intelligence to Address Electoral Integrity Issues” presents our project with Michelle Brown (National Democratic Institute), J. Andrew Harris (NYU Abu Dhabi) and Zach Warner (Cardiff) and showcases our efforts to detect voting fraud in even remote areas of developing countries. [youtube]

May 2020 Marcel Neunhoeffer (Mannheim) and I gave a quick overview over our working paper on the utility of differentially private synthetic data at the First OpenDP Community Meeting. [presentation, paper]