Chunk #48 — The Methods — How Do Classification and Regression Trees Work? — Model-Based Recursive Partitioning

Source

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

Embedded

yes

Text

A variant of recursive partitioning, that can also be a useful aid for visual data exploration, is model based recursive partitioning. Here the idea is to partition the feature space not such as to identify groups of subjects with similar values of the response variable, but groups of subjects with similar association patterns, e.g., between another predictor variable and the response.