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

Published in final edited form as: Mach Learn. 2005 Jun 8;61(1-3):71–103. doi: 10.1007/s10994-005-1123-6

Table 6.

Eleven point average precisions for classifiers trained to discriminate between two three-dimensional normal distributions. Training took place on one sample and testing on an independently generated sample from the same distributions. The weak learners for boosting are the three co-ordinate projections.

	N₁ mean	N₂ mean	Mahalanobis	Log-linear	Cmls	Lin-Ada	PAV-Ada
Set 1	(−1, −1, −1)	(2, 2, 2)	0.8556	0.7973	0.8235	0.8341	0.8357
Set 2	(0,0,0)	(1, 1, 1)	0.7250	0.5980	0.5961	0.5950	0.6107
Set 3	(0,0,0)	(1/2, 1/2, 1/2)	0.7014	0.5653	0.5606	0.5568	0.5624