Welcome to Marina Knight's Home Page

Marina Knight Home

Research Interests

Teaching

Publications

Collaborators

Marina Knight Senior RA School of Mathematics, University of Bristol, University Walk, Bristol. BS8 1TW Telephone: +44 (0)117 331 1575 Room 2.5, 67 St. Michael's Hill Email: marina.knight@bristol.ac.uk Brief Bio I completed a PhD in Statistics in 2006 at the School of Mathematics, University of Bristol, and then held a post-doctoral research position sponsored by GCHQ also at Bristol, until early 2008. During my time at Bristol, I developed a research interest in (second generation) wavelet techniques for analysing irregularly sampled data. Work in this area led to improved precision in nonparametric regression estimation as well as to applications in the field of bioinformatics. I also pursued research in the field of spectral estimation for non-stationary time series and proposed a method for spectral estimation of locally stationary processes that feature missing observations. Real-life data that possess such characteristics arise in a vast of array of fields from transportation and energy to medicine. Motivated by this spectral estimation problem, I (and co-author) also proposed the first construction for second generation nondecimated wavelets. This approach proved to be not only central to obtaining a spectrum estimate at irregular time points, but also led to building an aggregate nonparametric regression estimate with significantly lower risk even than those obtained by adaptive techniques. From early 2008 until mid 2011, I took a career break to fit with my very young family, and only worked part-time as a medical statistician for NHS Blood and Transplant. Since mid 2011, I returned to academia as a post-doctoral researcher, working on the exciting EPSRC funded project Locally Stationary Energy Time series (LETS) with a focus on the modelling and forecasting of energy-related non-stationary time series. The project is in partnership with University of Lancaster and industry-based contributors such as EDF (Paris), Shell (UK), Garrad Hassan (UK) and Nuon (UK). Go to the Statistics Group home page. Go to the Mathematics home page.