Casper Wu

Bayesian Cart Models

esent advances in Bayesian modelling and computation for CART
(classification and regression tree) models. The main work has been
covered in stat395 in last semester. This time first I will briefly
introduce Bayesian CART for those who have not heard that, including 
prior specification and  Metropolis-Hasting algorithm. Then I will 
try to compare my prior specification with those already in the 
literatures. More details on MCMC will be discussed and some 
exploratory tools on diagnosing convergence will be shown. 
Extensions to survival tree and introduction of random splitting 
level will be covered.