Decision Trees & Ensembles by Hand: Entropy, Information Gain, Random Forests & Boosting
Build a credit-default tree cell by cell: entropy and Gini, information gain with COUNTIFS, the best numeric cut-point, a two-level tree as nested IF, overfitting and pruning, a random-forest vote with out-of-bag rows, gradient boosting with residuals, and feature importance, in 8 graded labs.
The first lesson is free — no account needed.
About this course
A decision trees and ensembles course that builds a credit-default model for one fictional lender cell by cell in a spreadsheet, for data science students, analysts and anyone who wants to see inside tree-based models. You measure how mixed a node is with entropy and Gini impurity, score candidate splits by information gain with COUNTIFS, let the data choose the best numeric cut-point, grow a two-level tree and write it as one nested IF, compare training and test accuracy by depth and prune with cost-complexity, combine five trees into a random forest with a majority vote and out-of-bag rows, run three rounds of gradient boosting on the residuals with a learning rate and turn impurity decreases into feature importances with a chart. No coding required. The first lab is free.
Lessons
- Entropy & Gini: How Mixed Is a Node?Free
- Information Gain: Scoring Candidate Splits with COUNTIFS Pro
- The Best Numeric Threshold: Trying Every Cut-Point Pro
- Growing a Two-Level Tree: Leaves & Nested IF Predictions Pro
- Overfitting & Pruning: Train vs Test Accuracy by Depth Pro
- Bagging & Random Forests: Majority Vote and Out-of-Bag Rows Pro
- Gradient Boosting by Hand: Residuals, Stumps & a Learning Rate Pro
- Feature Importance from Impurity Decrease Pro