Hello there,

I have problems understanding how a dirichlet process is in fact a stochastic process; a random variable which changes over t \in T. What is the T here, is it the partitions of the sample space which can grow to infinity?

I would appreciate any explanations :) Thanks

| cite | improve this question | | | | |

Could try good old Wikipedia on this...

| cite | improve this answer | | | | |
  • 2
    $\begingroup$ If Wikipedia's page is the standard explanation, then this is a fantastic question. In fact, I would love to see a tutorial on DPs for the mathematician that has mastered the basics of measure-theoretic probability. $\endgroup$ – Neil Toronto Apr 21 '11 at 15:43
  • 1
    $\begingroup$ @Neil, I have no idea whether you formulated this request for real or as a jest but if you did ask seriously, a starting point could be the seminal paper Prior distributions on spaces of probability measures by T.S. Ferguson (Annals of Statistics, 1974). For a diversified and more recent list of references, see mlg.eng.cam.ac.uk/porbanz/talks/npb-tutorial.html $\endgroup$ – Did Aug 11 '11 at 23:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.