liveBook · Manning

Bootstrap sample

As explained in Timeless Algorithms, The Foundational Ideas

A dataset formed by drawing observations with replacement from an original dataset.

A bootstrap sample is formed by drawing from an original dataset with replacement. For a dataset of N observations, draw N times; an observation may appear more than once, while others may be left out. section 9.2.1 section 9.3.3

In bagging, each sample trains a model whose predictions are later aggregated; bootstrap sampling is the first step, not another name for the whole procedure. In a random forest, a new sample is drawn for each tree, alongside random feature selection at each split. section 9.2.1 section 9.3.3

How Timeless Algorithms, The Foundational Ideas covers it

How deep it goes
Defines N draws with replacement and explains that samples can contain duplicates and omit observations.
What it assumes
Assumes an original dataset of observations and defines sampling with replacement through repeated draws.
What you can do after
Apply bootstrap sampling to create training sets for trees in a random forest.
How it connects
Connects to bagging, decision trees, random forests, and out-of-bag evaluation.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.