MENTAL MODEL · M6537

Classification and Regression Tree

DecideHigh supportDecision Science

Included

Version 1.0.0 · Updated 2026-07-30

CORE DEFINITION

A decision tree algorithm proposed by Leo Breiman et al. in 1984. Core features: - Binary tree structure: each node produces only two branches - Classification and regression: can handle both classification and regression problems - Splitting criterion: uses Gini Index for classification, and least squares error for regression - Pruning mechanism: prevents overfitting through cost-complexity pruning. Scaffolding role: interpretable decisions, balancing accuracy and simplicity. Compared to "black box" models, the decision trees produced by CART are highly visual, with transparent and interpretable decision paths, especially suitable for scenarios where decision logic needs to be explained to non-technical personnel.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

Interpretable decisions, balancing accuracy and simplicity. Compared to "black box" models, the decision trees produced by CART are highly visual, with transparent and interpretable decision paths, especially suitable for scenarios where decision logic needs to be explained to non-technical personnel.

anchor

Anchor fast decisions

Based on "recursive binary splitting + impurity reduction". Select the optimal split using Gini (classification) or MSE (regression), recursively build the tree, and prune to control complexity and prevent overfitting.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Select the best feature/threshold to split. 2. Recursively build a binary tree. 3. Stop when stopping criteria are met. 4. Perform cost-complexity pruning. 5. Predict.

Check to track your progress (stored locally)

Learning progress0%

account_treeGenealogyexpand_more

menu_bookReferencesexpand_more

Source support: Explicit

link
en.wikipedia.orghttps://en.wikipedia.org/wiki/Decision_tree_learningZH · Explicit
verified

RELATED MODELS