Jack-Cherish/Machine-Learning

Provides end-to-end implementations with hand-coded algorithms (SMO for SVM, gradient ascent for logistic regression) alongside scikit-learn comparisons, enabling learners to understand mathematical foundations before using library abstractions. Each algorithm includes practical datasets—from handwritten digit recognition to contact matching and medical outcome prediction—paired with detailed tutorials covering both theoretical concepts and optimization techniques like stochastic gradient descent and kernel methods.

10,250 stars. No commits in the last 6 months.