« Counting | Main | n vs. n-1 »
September 20, 2005
Bayesian Inference
The textbook for this course takes the point of view of Classical Statistical Inference. There is an alternative that we can mention briefly, Bayesian Inference. If you look closely, you will see that people are Bayesians in their approach to life, where holding a subjective prior is not a problem.
Suppose A is observable, but we are really interested in B, which in more complex applications is an unknown parameter. Bayes' Theorem translates information on the probability of A given B into a statement about the probability of B given A.
We can then talk about the prior (unconditional) probablity of B and the posterior (conditional on A) probability of B.
Posted by bparke at September 20, 2005 09:51 PM