Avatar

Mark Webster

Data Scientist

Visformatics

About

Former Statistics postgraduate at University of Leeds (UK), now a Data Scientist at Visformatics. My current work is under wraps, but I aim to make this site a portfolio for older work and side projects. It’s currently under construction. I’m planning to migrate over my posts on my old blog. In the meantime, you can see them at the old Probability and Statistics blog, The Accidental Statistician.

Former posts:

  • Research Associate in Modelling, working in the department of Health Economics and Decision Modelling at the University of Sheffield.
  • Postdoctoral research fellow in Statistics,working on statistical modelling and calibration at the University of Leeds.
  • Postgraduate research student in Statistics, working on Monte Carlo Methods, at the University of Leeds.

Postdoc supervisor: John Paul Gosling

PhD supervisors: Jochen Voss, Stuart Barber.

Interests

  • Probability and Statistics
  • Statistical Modelling

Education

  • PhD in Monte Carlo Methods (Statistics), 2017

    University of Leeds

  • MMath in Mathematics, 2011

    University of Leeds

Recent Posts

R has no consistent table class

The usual case R, in addition to being array-based, can also be table-based: it has a table class in the base language, data.frame. …

Decomposing general scores as simple expectations

Previously, we decomposed the Brier score into simple expectations. We saw that, unlike the posterior decomposition, where we get the …

Decomposing the Brier score as simple expectations

There are plenty of articles out there on decomposing the Brier score, but they’re usually done in notation that’s …

Projects

*

autodb

Automatic database normalisation for R data frames.

Codex

Statistical modelling of a card game’s forum tournament data.

Thesis

Convergence properties of Approximate Bayesian computation

Publications

Optimising Bowel Cancer Screening Phase 1: Optimising the cost effectiveness of repeated FIT screening and screening strategies combining bowel scope and FIT screening

Evaluating the cost-effectiveness of bowel cancer screening programmes in the UK.

Convergence properties of approximate Bayesian computation

We also introduce a new variant of ABC.

The rate of convergence for approximate Bayesian computation

Our results can be used to guide the choice of the tolerance parameter.

Nano-optical observation of cascade switching in a parallel superconducting nanowire single photon detector

We explicitly demonstrate the single- and multi-photon triggering regimes.