The computation undertaken by a learning system can be viewed as occurring at two distinct times,  training time and consultation time. Consultation time is the time between when an  object is presented to a system for an inference to be made and the time when the inference is completed. Training time is the time prior to consultation time during which the system makes inferences from training data in preparation for consultation time. Lazylearning refers to any machine learning process that defers the majority of computation to consultation time. Two typical examples of lazy learning are  instance-based learning and Lazy Bayesian Rules. Lazy learning stands in contrast to eager learning in which the majority of computation occurs at training time.

Lazy learning can be computationally advantageous when predictions using a single  training setwill only be made for few objects. This is because only the immediate sections of the instance space that are occupied by...