Reversible Jump MCMC III
In the previous posting in this series on RJMCMC, I created a Stata program for the following problem. We have two binomial samples y1=8, n1=20 and y2=16, n2=30, but we are not sure whether to model them as, M1: binomials with different probabilities π1 and π2 M2: binomials with a common probability ξ. For my […]
Reversible Jump MCMC II
Last time I outlined the derivation of a Metropolis-Hastings algorithm for RJMCMC and this time I want to implement it in Stata for a very simple problem. I have taken the problem from the paper in The American Statistician by Barker and Link that I referenced at the end of my last posting. We have two binomial […]
Reversible Jump MCMC
I want to introduce RJMCMC and demonstrate how it can be programmed in Stata. To understand this algorithm, you really do need to know a little about the basis of Metropolis-Hastings, so this week I will run through the ideas behind the equations. A bit of algebra I am afraid, but once it is out […]
Back to Google Flu Trends
It has been rather a long gap since my last posting but I am told that the re-direction virus has been cleared from the host’s server so I am going to return to the topic of fitting Bayesian models to flu epidemics using the Google Flu Trends data (https://en.wikipedia.org/wiki/Google_Flu_Trends). Google’s project created estimates of the number of cases of […]
Recent Comments