Talk:Bayesian probability

Logic

I do not see much logic in the way this article is arranged. The title is "Bayesian probability", and a section is called "Bayesian probability calculus". What the difference may be, I wonder. Furthermore, Bayes' formula is just a theorem in (general) probability (calculus) and not especially in Bayesian theory, and hence "Bayes" formula" is not the main article of the section called Bayesian probability calculus. Neither is Bayesian inference the main article to this section. Nijdam (talk) 16:24, 31 August 2010 (UTC)

Okay, it is much better now. Nijdam (talk) 14:07, 1 September 2010 (UTC)

Subjective probability

I do not know whether all subjective probability is termed Baysian. Nijdam (talk) 08:31, 11 October 2011 (UTC)

Interpretation

The excerpt: 'Bayesian probability interprets the concept of probability as "a probability p is..."' is controversial. Reasons:

• stylistically, it does not look well written (the citation does not fit well into the sentence citing it)
• the cited text was written by E.T. Jaynes. Other Bayesians, such as deFinetti strongly disagree with it, thus, it is not appropriate as a general characterization
• even Jaynes in his later works disagrees with it, writing, that the purpose of assigning probabilities is actually the "plausible reasoning", not the "representing the state of knowledge"

Thus, the general characterization should either be replaced by a truly general one, or, at least, it should be mentioned, that the characterization is not general, and applies only to a particular POV Ladislav Mecir (talk) 00:29, 27 December 2011 (UTC)

Bayesian Frequentists

In the modern world, Bayesian and frequentist "interpretations" aren't diametrically opposed, and there is no need to consider oneself "a Bayesian" or "a frequentist". A Bayesian surely will not dispute that prior information is often obtained, whether directly or indirectly, with input from frequencies, and will surely not claim anything is wrong with the frequentist approach in the absence of additional prior information. Likewise, modern frequentists will happily use Bayesian probability so long as the conditional distributions lead to an ability to perform repeatable trials when such conditions are in place. — Preceding unsigned comment added by 149.97.32.36 (talk) 00:15, 11 January 2012 (UTC)

Are they said to be "diametrically opposed" in the article? There is a difference between them. This article needs to be reasonably clear about what the various varieties of Bayeasian-type probabilities are (how they areinterpreted). How they are used (and how their use differs/compares to frequentist and Neyman arguments) is better placed in Bayesian inference or statistical inference, or possibly in probability interpretations, or in a new article specifically for such philosophical matters. It may already be there to some extent. Melcombe (talk) 09:11, 11 January 2012 (UTC)

Readability for non-experts

This article is a perfect example of something I run into all the time on Wikipedia: technical articles are written by experts in the field and lack an explanation that is accessible to non-experts. Consider adding a paragraph at the beginning which explains the concept in simple, easy-to-understand terms.

I know that technical folks are often loathe to do this, because they are afraid of losing accuracy if they try to simplify things. If technical terms can't be avoided, they need references so that a non-expert can at least look them up. The first paragraph alone of this article would require links for "evidential probabilities", "extension" (of logic), "reasoning", "propositions", "truth or falsity", "uncertain", "evaluate", and "probabilist". But generally it would be way more pleasant if Wikipedia article started simple and got more complex as you read down. — Preceding unsigned comment added by Kunstcleaver (talk • contribs) 15:40, 8 February 2012 (UTC)