Talk:Pólya urn model

Bibliography

Does anyone knows the original Pòlya work on this? It appears to have been done in twenties... — Preceding unsigned comment added by 131.175.28.134 (talk) 17:30, 15 December 2011 (UTC)[reply]

The use of italics in lieu of clarity in 5th paragraph

Does anyone find it successful? (I do not.) — Preceding unsigned comment added by 50.161.38.194 (talk) 03:48, 26 September 2014 (UTC)[reply]

I feel the same way. I have no idea what the paragraph is trying to say and find it highly confusing indeed! — Preceding unsigned comment added by 193.52.24.75 (talk) 13:19, 10 April 2015 (UTC)[reply]

Would someone kindly write the article?

This article is full of irrelevant nonsense that precedes any statement of what the Polya urn model is. If there even is such a statement in this article.

This is not the place to go into an extended discussion of all urn models, all urn models related to the Polya urn model, etc. It is the place to state exactly what the Polya urn model is. This has not been done, at least without a huge amount of verbiage obscuring any definition of the Polya urn model — if there actually is any such definition in this article.

It needs to be fully erased, and then for someone to replace it: someone who a) knows exactly what the Polya urn model is and b) understands how to write an encyclopedia article. (I am not that person, since I don't know what the Polya urn model is.Daqu (talk) 15:51, 20 September 2015 (UTC)[reply]

I added the stuff about the Moran Model of population genetics. I would defend it being there. But you are right that a basic discussion of the properties of the basic Polya Urn Model is needed. My understanding is that in it, with two colors, the proportions of balls with one color asymptotically approaches a uniform distribution on (0,1). I think that this goes back to Polya's paper. I wish I were competent to write the needed discussion of the stochastic process in the basic Polya Urn Model. It is needed. Felsenst (talk) 15:18, 22 January 2016 (UTC)[reply]

How is it possible to have a discrete uniform distribution on an infinite number of values? Madyno (talk) 20:30, 26 February 2017 (UTC)[reply]

It isn't, but it is possible for a sequence of discrete random variables to approach a uniform distribution under weak convergence. Anyway I do not believe the previous comment on the uniform distribution is correct unless perhaps the bag starts with one ball of each colour, which is not usually assumed. In general, starting with some known number of balls the distribution will approach a beta distribution. SymplectoJim (talk) 14:33, 14 March 2019 (UTC)[reply]

I suggest that the article is started from scratch, even with a short placeholder. The structure of the article could follow that of the Chinese restaurant process article. SymplectoJim (talk) 14:33, 14 March 2019 (UTC)[reply]