High-Dimensional Data Analysis by John Wright and Yi Ma

High-Dimensional Data Analysis with Low-Dimensional Models:
Principles, Computation, and Applications

John Wright and Yi Ma, Cambridge University Press

Basic Information

The book covers new mathematical and computational principles for high-dimensional data analysis (statistics and geometry), scalable optimization methods (convex and nonconvex), and important applications such as scientific imaging, wideband communications, face recognition, 3D vision, and deep networks. This book is to be used as an introductory graduate textbook for the areas of data science, signal processing, optimization, and machine learning. It has been used as textbook for courses at UC Berkeley, Columbia University, and University of Michigan etc.

Download and Copyright

Download a pre-production copy: Book-Wright-Ma on Github or Book-Wright-Ma on Dropbox (version last updated on 2021-04-22).
A complete version of the manuscript was first released here online on December 06, 2020. The latest version reflects small changes made to the original.

Copyright statement: This material will be published by Cambridge University Press as "High-Dimensional Data Analysis with Low-Dimensional Models: Principles, Computation, and Applications" by John Wright and Yi Ma. This pre-publication version is free to view and download for personal use only, and is not for redistribution, re-sale or use in derivative works. Copyright ©Cambridge University Press 2018.

Citation (bibtex):

@book{Wright-Ma-2021,
author = {John Wright and Yi Ma},
title = {High-Dimensional Data Analysis with Low-Dimensional Models: Principles, Computation, and Applications},
publisher = {Cambridge University Press},
year = {2021}
}

Catalog Links

The book is currently under production. Hardcopy of the book will soon be available at: TBA

Resources and Errata

Related courses (future instructors: if you plan to use the textbook for your courses, please send us your course websites to share here):

UC Berkeley EECS 208: Computational Principles for High-dimensional Data Analysis, Fall 2021.
Columbia University ELEN 6876: Sparse and Low-dimensional Models for High-dimensional Data, Spring 2021.
University of Michigan EECS 559: Optimization Methods for Signal & Image Processing and Machine Learning, Spring 2021.

Some old course websites:

Reader's feedback: If you have any feedback, suggestions, or errors to report, please send us an email to: book.wright.ma@gmail.com.

Book errata: TBA.