Supervised Learning
A 22-module path through supervised learning, in study order from foundations to the frontier.
Modules
- Linear Regression and the Least Squares Objective
- The Normal Equations and Closed-Form Solution
- The Bias–Variance Tradeoff
- Ridge Regression and L2 Regularization
- Lasso Regression and L1 Sparsity
- Logistic Regression and the Sigmoid
- Maximum Likelihood and Cross-Entropy Loss
- The Perceptron Algorithm
- k-Nearest Neighbors
- Decision Trees and Recursive Partitioning
- Entropy and Information Gain
- Tree Pruning and Complexity Control
- Support Vector Machines and Maximum Margin
- Soft Margins and Hinge Loss
- Kernels and the Kernel Trick
- Naive Bayes Classifiers
- Linear Discriminant Analysis
- Bagging and the Bootstrap
- Random Forests
- AdaBoost and Exponential Loss
- Gradient Boosting as Functional Gradient Descent
- Stacking and Model Blending