Machine Learning — Core Concepts

What's Inside

• Linear regression — derived from first principles including the normal equation and gradient descent
• Logistic regression — full MLE derivation with the sigmoid, cost function, and gradient update
• Regularization — L1 vs L2 intuition, the geometry of sparsity, and the bias-variance tradeoff
• Decision trees — entropy, Gini impurity, information gain, with worked splitting examples
• Ensembles — bagging, random forests, gradient boosting, and XGBoost's objective
• Support vector machines — the margin, the dual, and the kernel trick with a concrete example
• k-NN, Naive Bayes, and k-Means — the classics, with their failure modes clearly marked
• Model evaluation — precision, recall, ROC, PR curves, cross-validation done right
• Feature engineering — encoding, scaling, leakage, and the mistakes that sink real models

Who It's For

• Engineers preparing for ML interviews at FAANG-scale companies
• Students taking a first ML course who want a reference that goes deeper than the slides
• Working data scientists who know how to use scikit-learn but never derived what's underneath
• Anyone who prefers handwritten intuition over dense textbook prose

Frequently Asked

• Can I print it? Yes. The PDF is laid out for both A4 and US Letter, with clear margins.
• Do I need a math background? High-school calculus and basic linear algebra are enough. Everything else is explained in context.
• What about updates? Every buyer gets every future revision of this set forever, emailed automatically.
• Refunds? Digital products aren't refundable once downloaded, but if you hit a defect, email us and we'll fix it or refund.