I am a computational economist, and have been a data scientist since long before there were data scientists. I have worked on diverse problems which involve computation, data, and theory, such as estimating demand for differentiated products, forecasting, pricing, quantum cosmology, bioinformatics, and more. Currently an independent data science consultant, I use good judgment, humor, data, and math to help clients solve their problems and optimize their businesses. My academic background includes an A.B. in Physics from Princeton, an MSc and a PhD in Economics from UCL, and postdoctoral research at the University of Chicago.
My academic research focuses on Industrial Organization, Econometrics, and Computational Economics, especially how to estimate demand for differentiated products. The papers from my thesis are available on SSRN. Much of this work concentrates on the benchmark Berry, Levinsohn, and Pakes (1995, 2004) model, including its finite sample properties and how to approximate high dimensional integrals both accurately and efficiently. Failure to approximate integrals accurately will cause the solver to stop in local optima and under-estimate standard errors.
In another paper, I evaluate the merger between Morrisons and Safeway in the UK by constructing a structural model of geographic demand for groceries.