Kiru Lab / Track
Classical Machine Learning
Fit, generalize, fail. Every failure mode of a large model appears here first, in miniature.
Classical machine learning is where the field's vocabulary was set: features, fitting, overfitting, bias, variance, generalization. Working through it on models small enough to hold in your head is the cheapest way to earn intuitions that transfer directly to systems too large to inspect. A linear model that memorizes its training set and a language model that regurgitates training data are the same phenomenon at different scales.
What does it mean for a model to have learned something rather than memorized it?
By the end you can
- Implement linear and logistic regression from scratch
- Implement gradient descent and diagnose it from its loss curve
- Explain overfitting in terms of capacity and data
- Design an evaluation split that cannot leak
- Read a decision boundary as a statement about a model's inductive bias
Module 1
Learning From Data
What a model is, how fitting works, and gradient descent built by hand.
Module 2
Generalization and Its Failures
Overfitting, capacity, leakage — the honest evaluation of a model.
Module 3
Classification and Decision Boundaries
Logistic regression, trees, and what a boundary tells you about a model's assumptions.