Bagging is an acronym for “Bootstrap Aggregation.” It is a machine learning ensemble method used to improve machine learning algorithms’ accuracy by reducing variance and diminishing overfitting.
Bagging is a parallel algorithm that allows training multiple algorithms simultaneously. It becomes possible by creating sub-samples of data. Bagging generates additional data for training from the dataset. This is achieved by random sampling with replacement from the original data.
Bagging is used, for example, in the random forest algorithms for combining the predictions generated by every single tree.