Implementing MCMC - Hamiltonian Monte Carlo

21 January 2021 — Written by Tom

#algorithms#bayesian statistics

This post is about Hamiltonian Monte Carlo, an MCMC algorithm that builds on the Metropolis algorithm, but uses information about the geometry of the posterior to make better proposals.

Implementing MCMC - the Metropolis algorithm

19 January 2021 — Written by Tom

#algorithms#bayesian statistics

I'm a big fan of probabilistic modelling and Bayesian inference. In fact at the time of writing that's the only topic I've written about on this blog so perhaps that's blindingly obvious...

Bayesian billiards

6 December 2020 — Written by Tom

#bayesian statistics#interactive

In his seminal paper "An Essay towards solving a Problem in the Doctrine of Chances", Thomas Bayes introduced a thought experiment involving six balls thrown randomly onto a billiards table.

Election Modelling - Part 3

9 December 2019 — Written by Tom

#bayesian statistics#politics

This is the third and final post in a series on election modelling; specifically multi-level regression with poststratification (MRP) and its successful use by YouGov in the 2017 general election...

Election Modelling - Part 2

11 November 2019 — Written by Tom

#bayesian statistics#politics#stan

This is the second post in a series on election modelling; specifically multi-level regression with poststratification (MRP) and its successful use by YouGov in the 2017 general election...

Older posts→