Probabilistic Machine Learning: An Introduction (Adaptive Computation and Machine Learning) ISBN-978-0262046824

This book provides a thorough and current overview of machine learning, including deep learning, using Bayesian decision theory and probabilistic modeling as a unifying framework. Mathematical underpinnings (such as linear algebra and optimization), fundamental supervised learning (such as logistic and linear regression and deep neural networks), and more complex subjects (such as transfer learning and unsupervised learning) are all covered in the book. Students can apply what they have learned through end-of-chapter tasks, and notation is covered in an appendix.