chapter  24
Evaluation of Classification Methods
ByNele Verbiest, Karel Vermeulen, Ankur Teredesai
Pages 24

First we discuss some generical validation schemes for evaluating classification models. Given a dataset and a classification model, we describe how to set up the data to use it effectively for training

and then validation or testing. Different schemes are studied, among which are cross validation schemes and bootstrap models. Our aim is to elucidate how choice of a scheme should take into account the bias, variance, and time complexity of the model.