Deep Learning: A Visual Approach
A friendly and complete guide to deep learning.
No prerequisites! Jump in and discover how deep learning works for yourself!
About the Book
A book for programmers, scientists, artists, engineers, educators, musicians, physicians, and anyone else who wants to understand and use deep learning. Our principles are clear explanations, over 1000 professional-grade illustrations, and no math (except for some addition and multiplication). The ideas are applicable to any computer, programming language, and library you want to use. You'll know how to design, build, and use existing and original DL systems, and how to make them work for you. The author is Andrew Glassner, a Principal Research Scientist at Weta Digital, where he uses computer graphics and deep learning to help artists create amazing, wonderful visuals for TV, films, and more.
Get Started Now
Just click on the cover to buy the book and start reading! You can also click here.
YouTube Overview!
I gave a 3.5 hour course based on these books at SIGGRAPH 2018 in Vancouver, BC in August 2018. They recorded the whole thing. You can watch the video on YouTube! Enjoy the intro music, or jump to the start of the video at about 1:40.
Friendly Writing
The book uses the same friendly and lucid tone that thousands of readers have enjoyed in my other books, papers, and my computer graphics column.
Enthusiastically Illustrated
Good illustrations can share some ideas better than words. The book contains over 1000 expertly conceived and executed images. Visual thinkers, rejoice!
Language and Library Independent
Except for two practical chapters based on Python libraries, nothing is tied to any particular language or library. We're all about the ideas, which apply to whatever system you want to use (including your own!).
No Prerequisites
If you can multiply and know how to write "Hello World" in any computer language, you're ready for this book. Nothing else is assumed, and everything is included. If you want to get the most out of the two practical chapters, a bit of Python knowledge will go a long way.
No Math
There's no math! We do everything with straightforward discussions and examples, and tons of images. There is literally no math other than plus, minus, times, and divide in the whole book.
Jupyter Notebooks Included
For those into Python, we include Jupyter notebooks for the practical chapters on scikit-learn and Keras, and also give you the notebooks to make every computer-generated figure in the book.
Table of Contents
- 1 Introduction to Machine Learning
- 2 Statistics
- 3 Probability
- 4 Bayes' Rule
- 5 Curves And Surfaces
- 6 Information Theory
- 7 Classification
- 8 Training And Testing
- 9 Overfitting And Underfitting
- 10 Neurons
- 11 Learning And Reasoning
- 12 Data Preparation
- 13 Classifiers
- 14 Ensembles
- 15 Scikit-Learn
- 16 Feed Forward Networks
- 17 Activation Functions
- 18 Backpropagation
- 19 Optimizers
First Edition Resources
This is an extensively revised and updated edition of my book. The first edition was e-book only, and is no longer available. But the figures and notebooks from that edition are still available for you to download for free. Get them here.