I followed the link from commenter Lemmus and encountered this list of resources on Bayesian reasoning. These were all fine, but they didn’t look like the Bayesian statistics that I am familiar with. (For example, the wikipedia link defines “A Bayes estimator is a statistical estimator that minimizes the average risk.” Which doesn’t have anything to do with anything that I do!) And the examples are all discrete–which makes sense, that’s a good way to start–but I think it gives a misleading perspective of Bayesian data analysis as a statistical method.
So I thought I’d post Chapter 1 from Bayesian Data Analysis so people could see a practicing statistician’s perspective. (Here’s the table of contents and front matter of the book, to place it in context.)
“Bayesian reasoning” and “Bayesian data analysis” have (almost) the same name, but they’re different things.
P.S. The distinction arises in other fields too. For example, a book on “thinking like a lawyer” will be different than a book on the practice of law, a book on “the engineering way of life” might not be very descriptive of what engineers actually do, and so forth.
P.P.S. If you’re coming from statistics or econometrics, and you want to see how Bayesian inference generalizes classical least squares and maximum likelihood, then Chapter 18 from my book with Jennifer may be the best place to start.
The link to Chapter 1 from BDA presently doesn't work.
Fixed; thanks.
The wikipedia entry is absolutely ludicrous as a description of Bayesian statistics.
But I have to point out that neither of us, nor anybody else who knows better has corrected it, so perhaps the blame is (partially) on us.
Oops,
seems I linked to the "Bayes" disambiguation page (the one you also link to) rather than the one I meant (now fixed, thanks for bringing it to my attention). I hadn't even read the "Bayes estimator" article and am completely ignorant about the topic. (Have not read your pdf yet.)
How do you feel about putting chapters or whole books (http://www.gaussianprocess.org/gpml/chapters/) online? What are the disadvantages for authors of a book to do so? I can certainly think of very many advantages for putting books online available.
Many, many thanks for putting chapter 18 online. As someone transitioning from econometrics to statistics, this is very useful.