Linear Models

← Back to Classical ML Algorithms

Models that make predictions using a linear combination of input features. Simple, interpretable, and fast — often the first model to try.

Key Properties

• L1 Regularization (Lasso)
• L2 Regularization (Ridge)

Types

• Linear Regression — continuous output, minimize MSE
• Logistic Regression — binary classification, sigmoid activation, minimize cross-entropy
• Regularized variants — L1 (Lasso, sparsity), L2 (Ridge, shrinkage), Elastic Net (both)

Related

• Support Vector Machines (linear with margin maximization)
• Neural Network Fundamentals (generalization of linear models)

---

ml linear-models regression classification