Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Deep learning
Goodfellow I., Bengio Y., Courville A., The MIT Press, Cambridge, MA, 2016. 800 pp. Type: Book (978-0-262035-61-3)
Date Reviewed: Aug 16 2017

Deep learning (DL) is an area of artificial intelligence (AI) that carries out machine learning in terms of a hierarchy of concepts. DL enables a machine to build complex concepts by combining simpler ones, and thus relieves a human operator from the burden of entering a huge volume of input training data to describe the world of the machine. Including its earlier forms (or names), DL has been present for more than 70 years; however, recent progress in hardware and software and the abundance of training data have made the entire technology world start talking about DL’s unprecedented growth and success. Nowadays, tasks such as image and speech understanding, necessary for applications ranging from web searching to self-driving vehicles, are or can potentially be the best fit for DL. DL could shift whole industries, and it is expected to revolutionize many sectors of the economy.

Goodfellow et al., some of the most innovative researchers in this field, chose the right time to compile mature and promising knowledge into this excellent book; it is projected to be the definitive book on DL for years to come.

The book consists of three parts. This first provides background on applied mathematics related to linear algebra and probability, and also describes some very basic concepts of machine learning such as overfitting, underfitting, Bayesian statistics, and (un)supervised learning. The second part focuses on the most established algorithms related to deep networks; in particular, it describes deep feedforward networks, regularization, convolutional networks, and recurrent networks, along with practical issues and applications pertaining to these subjects. The third part is about more speculative topics such as linear factor models, autoencoders, representation learning, approximate inference, and deep generative models.

This balanced mix of mathematical and algorithmic formalisms is appropriate for both undergraduate and graduate students studying a variety of disciplines, including machine learning, web search, computer vision, and natural language understanding. Moreover, practitioners developing software projects related, for instance, to image or speech recognition can use this book as a clear and comprehensive entry point to the area. The book can also be used as a reference point for these subjects. The broad but at the same time thorough coverage of DL will make any reader absolutely love it, but for data scientists this is must-have and must-read book.

More reviews about this item: Amazon, Goodreads

Reviewer:  Dimitrios Katsaros Review #: CR145490 (1711-0714)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Learning (I.2.6 )
 
 
Classifier Design And Evaluation (I.5.2 ... )
 
 
Applications And Expert Systems (I.2.1 )
 
Would you recommend this review?
yes
no
Other reviews under "Learning": Date
Learning in parallel networks: simulating learning in a probabilistic system
Hinton G. (ed) BYTE 10(4): 265-273, 1985. Type: Article
Nov 1 1985
Macro-operators: a weak method for learning
Korf R. Artificial Intelligence 26(1): 35-77, 1985. Type: Article
Feb 1 1986
Inferring (mal) rules from pupils’ protocols
Sleeman D.  Progress in artificial intelligence (, Orsay, France,391985. Type: Proceedings
Dec 1 1985
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy