"Deep Learning - A Visual Approach" by Andrew Glassner
This is the official repo for my book from No Starch Press.
Ordering the book
My book is called Deep Learning: A Visual Approach Click on the link to order it in physical or Ebook formats.
Free Bonus Chapters!
Three free bonus chapters! How to use scikit-learn for machine learning, and how to use Keras for deep learning. Free text, free notebooks, free figures, the whole thing! Just click here or click on the Bonus Chapters repo. The figures and notebooks are saved with all of the other figures and notebooks (see below).
Free Figures!
All the figures from my book, for free, in high-resolution
PNG format.
To help you search, there's a directory called Thumbnails
which
offers contact sheets of the figures, 20 per page.
All of these figures are released under the MIT license. This means you're free to use them any way you like, as long as you keep the copyright associated with them somehow. Use them for your classes, reports, papers, presentations, whatever you like!
You're not required to attribute me or the book if you use these images, but I'd appreciate it if you would.
Some figures include photographs. Many of these are by me, and I've given you permission to use them. All other photos are from Wikiart, Wikimedia, or Pixabay. The book provides a citation and URL to the source of each of these images. The first two sites state that their images are in the public domain. All images selected from Pixabay are labeled as released under the Creative Commons CC0 license, and explicitly state, "Free for commercial use. No attribution required."
Free Notebooks!
Jupyter notebooks for making many of the figures in the book.
Since the purpose of the notebooks was to make figures, rather than to serve as tutorials, they are only lightly commented, but they're meant to be readable. So I used longer but clearer variable names, and whenever I could I preferred clarity over most other concerns. This means that much of the code can be shortened, reorganized or otherwise refactored, and almost always it can be changed to be more compact, elegant, and faster. Feel free to dig in, optimize, convert to other languages, or otherwise play with the code.
All the notebooks are released under the MIT license. Informally, you're free to do pretty much anything with the code, including using it in your own projects, or even including it in commercial projects, as long as you keep my copyright along with the code. While I strove for accuracy and correctness, there is no warranty that the code is bug-free or fit for any purpose.
Some notebooks work with images. The images I used in the book are included with the notebooks. See the section below on Figures for details on their licensing, and see the book for the URL where each image may be found. All images without an explicit citation in the book are by the author, and are released under the MIT license.
Errata
A book of this size will inevitably have errors.
For each error I'm aware of, I'll update the appropriate figure(s)
and/or notebook(s), and then put a description of the error
(along with a credit to the person who found it) in a plain-text
file in the Errata
folder.