Arnold Pimentel Salas

DPhil Student, Engineering, Machine Learning Research Group

Arnold earned a BSc in Economics from the London School of Economics and Political Science, and a MSc in Finance from the Toulouse School of Economics (TSE). After completing his Master's degree, he worked as a postgraduate researcher in financial econometrics at TSE for two years, focusing on finite-sample efficient covariance matrix estimation for vast-dimensional asset portfolios. Then, he went on to work in the City for two years, as a Quantitative Researcher at Merrill Lynch, where he spent most of his time devising and implementing cross-asset quantitative trading strategies. Currently, he's doing a DPhil in Machine Learning in the Department of Engineering Science, with Prof Stephen Roberts and Dr Michael Osborne. His research project consists in developing novel online Bayesian algorithms to predict time series that are subject to changing population distributions, with a focus on financial data.


Working Paper

Rizvi, S., van Heerden, E., Salas, A., Nyikosa, F., Roberts, S., Osborne, M. and Rodriguez, E. (2017). Identifying sources of discrimination risk in the life cycle of machine intelligence applications under new European Union regs.
Salas, A., Roberts, S. and Osborne, M.A. (2015). A variational Bayesian state-space approach to online passive-aggressive regression.