Why not draw $b < n$ rows of the dataset, and fit an estimator on this subsample? Why is it important that the subsample have the same number of rows as the original dataset?
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
It is not necessary. Bagging is convenient when you want each of the replicate data sets to have the same size as the original data set. Otherwise you will have to determine the sizes of teach of these subsets. See this paper.
