Chunk #66 — The Methods — How Do Ensemble Methods Work? — Random Forests

Source

An introduction to recursive partitioning: rationale, application, and characteristics of classification and regression trees, bagging, and random forests.

Embedded

yes

Text

The reason why even suboptimal splits in weaker predictor variables can often improve the prediction accuracy of an ensemble is that the split selection process in regular classification trees is only locally optimal in each node: A variable and cutpoint are chosen with respect to the impurity reduction they can achieve in a given node defined by all previous splits, but regardless of all splits yet to come.