Haziq Jamil
Research student in statistics

Office: COL 2.0


I am a third-year PhD student in the Social Statistics group at the London School of Economics & Political Science. I am working in the area of statistical theory, methods and computation, and its application towards the social sciences. My supervisors are Dr. Wicher Bergsma and Prof. Irini Moustaki.

Together with Wicher, I am exploring the use of objective priors which we call I-priors in a function space setting in order to estimate linear models. By working with reproducing kernel Hilbert spaces, it would seem that linear models would benefit from these priors, especially in situations when maximum likelihood methods would tend to overfit. We have developed the I-prior methodology to fit various types of regression models including multilevel (a.k.a. mixed-effects) models and smoothing models. Currently, the work for my thesis is three-fold: 1) Computational methods for I-prior regression; 2) Bayesian variable selection using I-priors; and 3) estimation of covariance matrices using I-priors.

Previously, I worked at the Centre of Science & Technology, Research & Development (CSTRAD), at the Ministry of Defence, Brunei, where my primary task was to provide analytical insight and decision support to strategic acquisition projects. I graduated from Warwick University, completing the 4-year MMORSE degree in 2010. As my final year project, I used Bradley-Terry models applied to English Premier League football data. This project was supervised by Prof. David Firth.

Research interests

My work

Other things