Live AI/ML Sessions | Outside College & Office Hours | Session Recordings on LMS | Learn Around Your Schedule | Built for Students & Working Professionals
Live AI/ML Sessions | Outside College & Office Hours | Session Recordings on LMS | Learn Around Your Schedule | Built for Students & Working Professionals

cloudsandai

Decision Tree Classification: An Interactive Lab

Explore how decision trees split data into interpretable rules, compare Gini impurity with entropy, and see how tree complexity shapes nonlinear decision regions.

Back to Resources

Section A

Explore the dataset

Classify whether a student passes using study hours and practice tests. Passing depends on both features being high, creating a nonlinear corner-shaped region. Incomplete rows are ignored.

Study hoursPractice testsOutcomeAction

16 complete observations

Section C

Decision regions

Unlike logistic regression’s straight boundary, a decision tree makes axis-aligned splits that form rectangular regions.

Study 1 hours, 1 practice tests, actual class 0Study 1 hours, 2 practice tests, actual class 0Study 2 hours, 1 practice tests, actual class 0Study 2 hours, 2 practice tests, actual class 0Study 1 hours, 5 practice tests, actual class 0Study 1 hours, 6 practice tests, actual class 0Study 2 hours, 5 practice tests, actual class 0Study 2 hours, 6 practice tests, actual class 0Study 5 hours, 1 practice tests, actual class 0Study 5 hours, 2 practice tests, actual class 0Study 6 hours, 1 practice tests, actual class 0Study 6 hours, 2 practice tests, actual class 0Study 5 hours, 5 practice tests, actual class 1Study 5 hours, 6 practice tests, actual class 1Study 6 hours, 5 practice tests, actual class 1Study 6 hours, 6 practice tests, actual class 1Study hoursPractice tests
Fail · class 0 Pass · class 1 Tree split

Section D

Learn the rules

Each root-to-leaf path becomes an interpretable IF–THEN rule.

IF Study hours ≤ 3.50 THEN predict fail (class 0) · 8 samples

IF Study hours > 3.50 AND Practice tests ≤ 3.50 THEN predict fail (class 0) · 4 samples

IF Study hours > 3.50 AND Practice tests > 3.50 THEN predict pass (class 1) · 4 samples

Evaluate the tree

Accuracy

100.0%

Precision

100.0%

Recall

100.0%

Leaves

3

Confusion matrix · rows are actual, columns are predicted

Predicted failPredicted pass
Actual fail120
Actual pass04