Hidden Markov mixtures of regression

It took the RSS feed of Bayesian Analysis to disappear from my screen—because the Bayesian Analysis 4(4) issue was completed—for me to spot this very nice paper by Matthew A. Taddy and Athanasios Kottas on Markov switching regression models. It reminds me of earlier papers of mine’s with Monica Billio and Alain Monfort, and with Merrilee Hurn and Ana Justel, on Markov switching and mixtures of regression, respectively. At that time, with Merrilee, we had in mind to extend mixtures of regressions to generalised linear mixtures of generalised linear models but never found the opportunity to concretise the model. The current paper goes much farther by using mixtures of Dirichlet priors, thus giving a semi-parametric flavour to the mixture of regressions. There is also an interesting application to fishery management.

This issue also includes an emotional postnote by Brad Carlin, who is now stepping down from being the Bayesian Analysis Editor-in-chief. Brad unreservedly deserves thanks for mentoring Bayesian Analysis towards a wider audience and a stronger requirement on the papers being published in the journal. I think Bayesian Analysis now is a mainstream journal rather than the emanation of a society, albeit as exciting as ISBA! The electronic format adopted by Bayesian Analysis should be exploited further towards forums and on-line discussions of all papers, rather than singling out one paper by issue, and I am glad Brad agrees on this possible change of editorial policy. All the best to the new Editor-in-chief, Herbie Lee!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: