Machine Learning with R Cookbook - Second Edition

How it works...

Bagging is derived from the name Bootstrap aggregating, which is a stable, accurate, and easy to implement model for data classification and regression. The definition of bagging is as follows: given a training dataset of size n, bagging performs Bootstrap sampling and generates m new training sets, Di, each of size n. Finally, we can fit m Bootstrap samples to m models and combine the result by averaging the output (for regression) or voting (for classification):

An illustration of the bagging method

The advantage of using bagging is that it is a powerful learning method which is easy to understand and implement. However, ...