Chunk #38 — The Methods — How Do Classification and Regression Trees Work? — Prediction and Interpretation of Classification and Regression Trees

Source

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

Embedded

yes

Text

For classification problems it is also possible to predict an estimate of the class probabilities from the relative frequencies of each class in the terminal nodes. In our example, the predicted probabilities for answering “yes” to the intention to smoke question would thus be approximately 15%, 30% and 65% in the three groups – which may preserve more information than the majority vote that merely assigns the class with a relative frequency of > 50% as the prediction.