Shown below are average cross-validation balanced error rates of UCI datasets (with 10 90:10 train validation splits)

Dataset	Nearest means	Naive bayes	Kernel nearest means (poly degree 2)	Perceptron
Breast cancer	15.9%	7.6%	21.5%	20.6%
Climate	15.1%	20.9%	14.7%	50%
Qsar	33.4%	23.4%	34.7%
Ionosphere	27.9%	31.5%	30.7%	18.9%
Micromass	26.4%	22.4%	37.3%
Hill valley	46.8%	47.3%	48.4%