Classical ML & Ensembles
The algorithms that still win on tabular data
Trees, forests, boosting, SVMs, and clustering — the workhorse algorithms that beat deep learning on most real business data.
Intermediate3 modules · 10 lessons~78 min715 XPAligned with Google MLCC · scikit-learnStart with: ML Foundations
Progress0/10
Module
Trees & Forests
If-else questions, asked at scale
How a Tree Decides8 min · 70 XPHow a Tree Decides
Random Forests: Wisdom of the Crowd8 min · 75 XPRandom Forests: Wisdom of the Crowd
Gradient Boosting: Serial Self-Correction9 min · 80 XPGradient Boosting: Serial Self-Correction
The Tabular Truth7 min · 65 XPThe Tabular Truth
Module
Boundaries & Margins
SVMs, nearest neighbors, and the baselines that set the bar
SVMs: The Widest Street8 min · 75 XPSVMs: The Widest Street
k-NN: Learning by Looking Around7 min · 65 XPk-NN: Learning by Looking Around
The Baseline You Must Beat7 min · 65 XPThe Baseline You Must Beat
Module
Unsupervised Learning
Structure without labels — clustering and compression
K-Means and Its Blind Spots8 min · 75 XPK-Means and Its Blind Spots
Hierarchical Clustering7 min · 65 XPHierarchical Clustering
Dimensionality Reduction9 min · 80 XPDimensionality Reduction
Capstone project
Algorithm bake-off
Same dataset, three algorithms — e.g., logistic regression, decision tree, SVM. Tune each fairly, then explain why the winner won in terms of the data's structure.
Deliverable: A notebook with comparable metrics and a written explanation of the winner's advantage.
Milestones
Finish every lesson to earn the Classical Machine Learning Certificate.