|Phone (external)||+44(0)20 3108 3227|
I am currently a research associate, investigating new methods to approximate the likelihood function in the analysis of large spatial datasets. I completed my PhD at University College London Statistical Sciences Dpt in November 2017, after obtaining my MsC degree from Ecole Centrale de Nantes in France in September 2014, where I specialized in computer sciences.
During my PhD, I researched new non-stationary models and inference methods based on spectral domain quasi-likelihood. I also contributed to developping the asymptotics theory of a new de-biased spectral domain estimator for stationary time series (see publications below).
In my free time I have recently developed an interest for deep learning methods and AI.
Time series, random fields, correlation theory in time and space, non-stationary models and inference,
computer-efficient algorithms, likelihood methods, likelihood approximations, classification.
- A. P. Guillaumin, A. M. Sykulski, S. C. Olhede, J. J. Early, J. M. Lilly (2017). Non-stationary
modulated time series with applications to ocean flow measurements. Journal of Time Series
- A. M. Sykulski, S. C. Olhede, J. M. Lilly, A. P. Guillaumin, J. J. Early. The de-biased Whittle likelihood for
second-order stationary stochastic processes. Biometrika, conditionally accepted.