Bayes Taster and BUGS, JAGS, Stan, PyMC3 webinars
This article is originally published at https://robertgrantstats.wordpress.com
The Bayes Taster webinar is back next week, 18 February. It’s a one hour introduction to Bayesian analysis that focuses on giving you an intuitive grasp of what Bayes is, how it differs to what you might already know from statistics or machine learning, and what it is most useful for.
I designed this taster webinar because there are many ways to get in-depth, technical knowledge of the subject, but that is a commitment of time and money. I think there are lots of people out there who want to just learn a little more first, before taking that decision. If you are managing a data science team, or considering Bayesian methods as a new skill, this probably means you!
The Bayes Taster costs a mere 5 pounds. Like, a coffee and a slice of cake. Go on, you will, you will.
Then, I am going to give three taster webinars looking at particular software options. I’m not considering point ‘n’ click interfaces or preset commands inside other software, even though some of these are pretty good. I’m looking at the serious options — probabilistic programing languages — which let you code up your own model in your own way. They are harder to learn, but they keep on being useful, however complex your analysis gets. All of these are open-source and free.
First up is a one-hour taster looking at BUGS and JAGS: the original probabilistic programming languages, or at least model scripting languages. They’re going strong, and a lot of people are still happily using WinBUGS as a standalone package. There are interfaces from R, Python and even Stata. That’s on 1 March and costs £30.
On 4 March, you get another one-hour taster on Stan, a newer probabilistic programming language that uses an improved algorithm and solves a load of BUGS/JAGS problems. However, it doesn’t beat BUGS/JAGS at everything, and I’ll look at that too.
Finally, on 8 March, another one-hour taster looking at PyMC3, a Python package for Bayesian inference that contains the best bits of the others we’ve looked at so far.
You can also get a ticket to all of them for £65, which basically means you get one of the software webinars for free. At the end of this series, you’ll know how you could be using Bayesian modelling, how difficult it might be in your context, and what software option you would be most comfortable with.
And, while you’re here, there are more BayesCamp sessions coming up on things like data visualisation, Stata, clinical audit, meta-analysis… take a look at the courses page on the website.
Please visit source website for post related comments.