Encyclopedia of Machine Learning

Editors: Claude Sammut, Geoffrey I. Webb

Training Time

Reference work entry
DOI: https://doi.org/10.1007/978-0-387-30164-8_844

Definition

A learning algorithm is typically applied at two distinct times. Training time refers to the time when an algorithm is learning a model from  training data. Test time refers to the time when an algorithm is applying a learned model to make predictions.  Lazy learning susually blurs the distinction between these two times, deferring most learning until test time.

Copyright information

© Springer Science+Business Media, LLC 2011