Data Science with Julia by McNicholas Paul D.; Tait Peter;

Author:McNicholas, Paul D.; Tait, Peter;
Language: eng
Format: epub
Publisher: CRC Press LLC

5.2.2 K-Fold Cross-Validation

K-fold cross-validation partitions the training set into K (roughly) equal parts. This partitioning is often done so as to ensure that the yi are (roughly) representative of the training set within each partition — this is known as stratification, and it can also be used during the training-test split. In the context of cross-validation, stratification helps to ensure that each of the K partitions is in some sense representative of the training set. On each one of K iterations, cross-validation proceeds by using K − 1 of the folds to construct the learner and the remaining fold to compute the error. Then, after all iterations are completed, the K errors are combined to give the cross-validation error.

The choice of K in K-fold cross-validation can be regarded as finding a balance between variance and bias. The variance-bias tradeoff is well known in statistics and arises from the fact that, for an estimator of a parameter,

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.