Concepts

In teaching order. Each builds on the ones before it.

1. Descriptive statistics · 33 implementations · beginner

   Count, center, and spread of one numeric column.

2. Linear regression · 33 implementations · beginner

   Fit a straight line to data by least squares.

3. k-nearest neighbors · 29 implementations · beginner

   Classify a point by the majority vote of its closest neighbors.

4. k-means clustering · 30 implementations · intermediate

   Group points into k clusters around moving centroids.

5. Naive Bayes · 29 implementations · intermediate

   Classify by multiplying per-feature probabilities learned from counts.

6. Logistic regression · 30 implementations · intermediate

   Predict a yes-or-no outcome by fitting probabilities with gradient descent.

7. Decision tree · 29 implementations · intermediate

   Learn yes-or-no rules by repeatedly splitting the data where it separates best.

8. Principal component analysis · 33 implementations · advanced

   Find the few directions that capture most of the variation in many columns.

9. Cross-validation · 29 implementations · intermediate

   Choose a model's settings by testing it on every part of the training data in turn.