CART

As explained in Grokking AI Algorithms, Second Edition

CART is a decision-tree algorithm that constructs binary trees for classification and regression.

CART (Classification and Regression Tree) learns a binary decision tree by asking feature-based questions that separate examples into their categories. For diamond data, it might ask whether Carat is at least 0.5 or whether Price exceeds a threshold. section 8.17

A practical CART training loop works like this:

• Represent each decision as a question, then split examples into true and false groups. section 8.17
• Measure impurity with the Gini index, which reflects how mixed the labels are. section 8.17
• Test feature values, choose the split with the greatest information gain, and repeat recursively. section 8.17
• Stop when a split produces no gain and store the remaining label counts in a leaf node. section 8.17

Figure 8.19 A basic flow for building a decision tree

Use the flow diagram to follow CART from candidate questions through data splits to a trained tree. section 8.17

How Grokking AI Algorithms, Second Edition covers it

How deep it goes

Explains data structures, Gini uncertainty, information gain, candidate splits, recursion, and stopping, while leaving formal theory aside.

Examples

Uses Okay and Perfect diamond labels, Carat and Price questions, tree diagrams, and Python implementations.

What it assumes

Assumes familiarity with labeled examples and basic tree-like data structures while developing the algorithm incrementally.

What you can do after

Builds a basic decision tree from scratch using questions, Gini calculations, information gain, nodes, and recursion.

How it connects

Connects decision trees with entropy, Gini index, information gain, classification, and overfitting.

liveBookmanning.com

Read Grokking AI Algorithms, Second Edition on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Machine Learning Algorithms in DepthCART is a decision-tree learning algorithm that recursively partitions data using optimized split criteria.

Compare how 2 books explain CART →

Related concepts

• decision treeimplements · A decision tree is a branching structure that represents sequential tests or choices, with paths leading to outcomes, predictions, or actions.
• Gini impurityuses · Gini impurity is a measure of class-label heterogeneity in a set or decision-tree node, calculated from the proportions of its classes.
• divide-and-conquer approachuses · Divide and conquer is a problem-solving strategy that decomposes a problem into smaller parts, handles them separately, and combines or completes their results.
• recursionkind of · Recursion is a computational technique in which a definition or procedure refers to or invokes the same definition or procedure, directly or indirectly.

This explainer was generated by Manning's knowledge graph from the text of Grokking AI Algorithms, Second Edition. The section links point to the parts of the book it draws on.