Book Review : Deep Learning By Goodfellow, Bengio, and Courville
Who can explain deep learning better than Goodfellow, Bengio, and Courville? They cover deep learning in depth while still remaining comprehensible to those who have just started the subject. Deep Learning By Goodfellow, Bengio, and Courville is an amazing book, useful for students, professors and even practitioners in the same field. But before you start, you should be familiar with basic programming and calculus.
Who Are Goodfellow, Bengio, and Courville?
Ian Goodfellow is a research scientist who has a Ph.D. in machine learning, currently employed at Google.
Aaron Courville is an Assistant Professor at the University of Montreal.
Yoshua Bengio is a computer scientist, full Professor at the department of computer science at the University of Montreal.
The book is divided into three parts/sections.
Part I: Applied Math and Machine Learning Basics
The first part covers applied math and machine learning basics. This is great for people who have just started learning about machine learning and for those who want to go over the topic to refresh their memories. This first part gives the reader an introduction to probability theory, linear algebra, optimization, and it goes over some machine learning concepts.
This section goes over the different machine learning algorithms and other related concepts which aren’t just those relating to deep learning. Furthermore, it covers the basic concepts you need to build machines which are able to learn from data. Additionally, this section includes discussions on different topics such as machine learning algorithm capacity, overfitting, and underfitting. This section provides you with the context for deep learning in the grader machine learning field.
Part II: Modern Practical Deep Networks
The second part covers deep learning concepts, Artificial Neural networks(ANN) and types of ANN such as recursive neural network, deep multilayer perceptron (MLP), convolutional neural network (CNN), recurrent neural network. Also, this section focus on feedforward neural networks, the construction of them, and training. This part of the book targets those who have more knowledge about deep learning. It mentions the choices which need to be made to use deep learning to build accurate and useful predictive machine systems.
Part III: Deep Learning Research
The third part of this book is dedicated to deep learning research concepts. It includes topics which are relevant to researchers who want to push the field of deep learning even further. This part of the book covers topics such as approximate inference, autoencoders, and generative models.
The only problem with the book is that its more than 700 pages and has no practical problems. Its all about theory. In terms of theories, this is one of the best deep learning books you should be reading, so if you want to read about deep learning this a book you should purchase.
You will find a hardcover of the book here Deep Learning (Adaptive Computation and Machine Learning series)
An online version of the book with other lectures and exercises can be found on their official website Deep Learning (Adaptive Computation and Machine Learning series)
If you are interested in Artificial Intelligence, machine learning and deep learning, check these articles.