Quasi-Newton Method for Logistic Binary Classification
Implementing Logistic Regression with the BFGS quasi-Newton method and L-BFGS. Covers mathematical derivation of quasi-Newton conditions, BFGS update formulas, and Wolfe line search.
Tag: 机器学习
Logistic Regression Implementation
Logistic regression is a supervised classification model. This article implements binary logistic regression using gradient descent and Newton's method, deriving loss functions, gradients, and Hessian matrices with full Python code.
Multiple Linear Regression Implementation
Multiple linear regression implemented via the normal equation and gradient descent, with matrix calculus derivations and L2 regularization.