BART Procedure
PROC BART Features
PROC BART has the following basic features:
trains Bayesian additive regression trees (BART) models for normally distributed response data and probit models for binary response data
provides a STORE statement to create an analytic store for the model and saves it as a binary object in a data table
provides a FREQ statement for grouped analysis
provides an OUTPUT statement to produce a data table that contains predicted values and other observationwise statistics
provides a PARTITION statement to divide data into training data and test data roles
provides a TREEPRIOR statement to specify the BART prior parameters for the sum-of-trees ensemble
provides a GAUSSPRIOR statement to specify the BART prior parameters for a model that is fitted to normally distributed data
Because the BART procedure runs on SAS Cloud Analytic Services (CAS), it also does the following:
enables you to run on a cluster of machines that distribute the data and the computations
enables you to run in single-machine mode
exploits available cores and concurrent threads