Kevin P. Murphy, MIT Press, 2022

Probabilistic
Machine Learning

The probabilistic foundations of modern ML — from Bayes' theorem to deep networks to Gaussian processes. Every model is a statement about uncertainty.

Chapters

30+

Simulations

60+

Quizzes

Part I: Foundations

Chapters 2–3

Probability

Random variables, Bayes' rule, Gaussians, multivariate distributions, conjugate priors, exponential family.

Chapter 4

Statistics

Maximum likelihood, MAP estimation, regularization, Bayesian statistics, bias-variance tradeoff.

Part II: Linear Models

Chapters 10–11

Linear Models

Logistic regression, decision boundaries, linear regression, ridge, lasso, Bayesian linear regression.

Part III: Deep Neural Networks

Chapters 13–15

Deep Neural Networks

MLPs, backpropagation, CNNs, RNNs, attention, transformers, training and regularization.

Part IV: Nonparametric Models

Chapters 16–17

Nonparametric Methods

KNN, kernel density estimation, Gaussian processes, SVMs, kernel methods.

Part V: Beyond Supervised Learning

Chapters 20–21

Generative Models & Clustering

PCA, autoencoders, VAEs, K-means, GMMs, EM algorithm, spectral clustering.

ProbabilisticMachine Learning