. Author manuscript; available in PMC: 2018 Nov 16.

Published in final edited form as: Stat Comput. 2017 Jul 27;28(4):869–890. doi: 10.1007/s11222-017-9767-1

Table 4.

Prostate cancer data: Classification rate, area under the precision recall curve (AUPRC) and CPU run time in seconds (standard deviation in brackets) for BART-BMA, RF and BART. Methods with higher classification rate, AUPRC and lower CPU running time are more desirable. Elements in bold show the best method with respect to each of the criteria.

	BART BMA	RF	ERT	bart Machine	dbarts
Classification
Rate	0.79	0.71	0.75	0.71	0.69
AUPRC	0.68	0.67	0.60	0.65	0.60