Friday, February 11, 2005

Uninformative Priors

I came across an interesting article by Zhu and Lu talking about how non-informative priors can often be counter-intuitive. For instance suppose we are estimating parameter of a Bernoulli distribution. If we don't know anything, the most intuitive prior to use is uniform. However Jeffreys' "uninformative" prior doesn't look uniform at all:



This prior seems to favour values of theta close to 0 or 1, so how can it be uninformative?


The intuition they provide is the following: think of Maximum Likelihood estimator as the most uninformative estimate of parameters. Then we can rank informativeness of priors by how closely the Bayesian Posterior Mean estimate comes to the maximum likelihood estimate.

If we observed k heads out of n tosses, we get following estimates:

1. MLE: k/n
2. PM using Jeffrey's prior (k+1/2)/(n+1)
3. Laplace smoothing (Uniform) (k+1)/(n+2)
4. PM using Beta(a,b) prior (k+a)/(n+a+b)

You can see that using this informal criteria for informativeness, Jeffreys' prior is more uninformative than Uniform, and Beta(0,0) (the limiting distribution) is the least informative of all, because then Posterior Mean estimate will coincide with the MLE.

Resources:


6 comments:

Anonymous said...

To nit-pick, the Kass & Wasserman paper is 1995, not 1991, and Rob has it on his homepage,
http://www.stat.cmu.edu/~kass/papers/rules.pdf. It's still a good place to begin thinking about these and related topics...

Yaroslav Bulatov said...

Thanks for the note, fixed

draj said...
This comment has been removed by a blog administrator.
Miller said...

All these mathematical formulas, calculations and graphs scare me. I have never been friends with mathematics. This is too hard for me. We had a guy in our class, everyone called him a mathematical genius and he solved mathematical problems for me, and I wrote essays for him. I loved to write, my friends often turned to me for help in this matter, I helped them and at the same time worked on the service https://www.rush-my-essay.com/pay-for-essay/ and helped many more students, but for money. Writing turns out to be not only interesting for me but also brings good income as a student.

idadalton said...

The greatest gambling platform is this one.
When I play at Ratemycasino, I am really pleased because their website is extremely simple to use and play the games. I also really desire casino like this because they offer so many events where you can gain a lot of money.

John Fei said...

When faced with the daunting task of writing an essay, the "write my essay ai" service comes to the rescue. By leveraging the power of artificial intelligence, this service generates well-crafted essays based on your requirements. It eliminates the stress and time-consuming process of writing, allowing you to focus on other important tasks. The "write my essay AI" service is a reliable companion that ensures you receive top-notch written content without the hassle.