Skip to main content

Advertisement

Table 4 Results for the microarray example

From: Estimating misclassification error: a closer look at cross-validation based methods

Method n ē a ē N M S E SD(MSE) V A R SD(VAR) MSB B I A S SD(BIAS)
LOOCV 50 0.2363 0.2404 0.0056 0.0077 0 0 0.0056 0.0042 0.0746
BCVn 50 0.2363 0.1726 0.0104 0.0070 0.0037 0.0013 0.0067 −0.0636 0.0517
k CVn/2 50 0.2363 0.2418 0.0056 0.0074 0.0003 0.0001 0.0053 0.0055 0.0729
BCVn/2 50 0.2363 0.1736 0.0102 0.0069 0.0037 0.0011 0.0065 −0.0627 0.0511
k CV10 50 0.2363 0.2454 0.0057 0.0070 0.0007 0.0003 0.0050 0.0091 0.0704
BCV10 50 0.2363 0.1772 0.0098 0.0066 0.0037 0.0010 0.0061 −0.0591 0.0508
  1. e a = 1 / N i = 1 N e i , e N = 1 / N i = 1 N e Ri , M S B = 1 / N i = 1 N ( e Ri e i ) 2 .