Measure

Intrinsic error component (squared) of the bias-variance decomposition as defined by Kohavi and Wolpert in: R. Kohavi and D. Wolpert (1996), Bias plus variance decomposition for zero-one loss functions, in Proc. of the Thirteenth International Machine Learning Conference (ICML96) This quantity is a lower bound on the expected cost of any learning algorithm. It is the expected cost of the Bayes optimal classifier. Estimated using the classifier using the sub-sampled cross-validation procedure as specified in: Geoffrey I. Webb & Paul Conilione (2002), Estimating bias and variance from data , School of Computer Science and Software Engineering, Monash University, Australia

Minimum value | |

Maximum value | |

Unit | |

Optimization | Lower is better |