MENTAL MODEL · M7296

Binary Classification Tree

DecidemediumDecision Science

Included

Updated 2026-08-10

Loading revision record…

INTRODUCTION

English translation pending.

CORE DEFINITION

The binary classification tree, formalized in methods such as CART by Breiman and colleagues, recursively partitions data by asking one binary question at each node. Its core proposition is that a sequence of simple splits can approximate complex decision boundaries while remaining readable as rules. The key qualifiers are pruning and stability: an unpruned tree memorizes noise, and small changes in data can restructure the splits, so depth control and validation are part of the method rather than optional refinements.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

- Rule ordering: turn a complex judgment into ordered yes-or-no tests at each step. - Path reading: read a single path to extract one explicit and checkable decision rule. - Branch pruning: cut branches that fit noise rather than the underlying signal in the data.

anchor

Anchor fast decisions

Each split selects the feature and threshold that most reduces impurity in the child nodes. Repeating this greedily produces partitions that separate classes with few questions. Because every decision is a local comparison, the resulting path can be stated in words, which is what makes the model explainable where weighted sums of many features are not.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Choose a splitting criterion such as Gini impurity or entropy. 2. Split recursively until a stopping rule is reached. 3. Prune branches that do not improve validation performance. 4. Read the surviving paths and state them as rules.

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/Binary_classificationZH · Explicit
verified

RELATED MODELS