Shewchuk Jonathan Richard. Concise Machine Learning

Department of Electrical Engineering and Computer Sciences University of California at Berkeley, 2017. — 164 p.Linear Classifiers and Perceptrons
Perceptron Learning; Maximum Margin Classifiers
Soft-Margin Support Vector Machines; Features
Machine Learning Abstractions and Numerical Optimization
Decision Theory; Generative and Discriminative Models
Gaussian Discriminant Analysis, including QDA and LDA
Eigenvectors and the Anisotropic Multivariate Normal Distribution
Anisotropic Gaussians, Maximum Likelihood Estimation, QDA, and LDA
Regression, including Least-Squares Linear and Logistic Regression
More Regression; Newton’s Method; ROC Curves
Statistical Justifications; the Bias-Variance Decomposition
Shrinkage: Ridge Regression, Subset Selection, and Lasso
The Kernel Trick
Decision Trees
More Decision Trees, Ensemble Learning, and Random Forests
Neural Networks
Neurons: Variations on Neural Networks
Better Neural Network Training; Convolutional Neural Networks
Unsupervised Learning and Principal Components Analysis
The Singular Value Decomposition; Clustering
Spectral Graph Clustering
Learning Theory
Multiple Eigenvectors; Latent Factor Analysis; Nearest Neighbors
Faster Nearest Neighbors: Voronoi Diagrams and k-d Trees