This book is a short introduction to deep learning for readers with a STEM background. It aims at providing the necessary context to understand landmark AI models for image generation and language understanding.
François Fleuret is Full Professor and head of the Machine Learning group in the department of Computer Science at the University of Geneva where he holds the chair of machine learning. He received his PhD in Mathematics from INRIA and the University of Paris VI in 2000.
He is the inventor of several patents in the field of machine learning, and co-founder of Neural Concept SA, a company specializing in the development and commercialization of deep learning solutions for engineering design.
I did not get any of the math but sometimes it’s useful to be made aware of all the things you don’t understand about a topic, despite basically doing related work on it for years and years.
This was an excellent little book, very thoughtfully available for free in a PDF formatted to fit on a mobile device. A really good overview, naturally fairly light on detail, but still extremely impressive what the author managed to fit into those little chapters! 4 phone-screen-sized pages to tell you all about Transformers? No problem! Good work, and a great dive into the topic.
This book gives a good overview of the different aspects of deep learning starting with fundamentals building up to neural networks, Deep learning, model architectures and eventually applications. The math is sometimes hard to understand so you have to look at supplemental resources. In terms of coverage and references the book does very well though it does skip over some concepts such as RNNs and assumes some fore-knowledge of neural networks and deep learning.
This book is targeted towards the intermediate to advanced AI engineer. A small and mighty book that takes times to go through as the information density is is high and you have to read up the references and associated concepts not covered in detail. But it works well as a compact guidebook to the different concepts in the deep learning area.
This is a good quick review of Deep Learning ideas and techniques. It helps to have some background in this space as some parts assume knowledge of DL concepts. My favorite part was chapter 5, "Architectures", which described the different DL architectures. The diagrams were excellent and clear. It is a quick read.
The only issue is who is the target audience. For experts they already know everything in here. For novices it assumes too much background. I think it's targeted towards intermediate folks (such as myself) who are in the middle of learning DL. Overall, it's a good review of concepts and introduction to some new concepts I did not know about (Transformers, Attention, etc).
One other thing: the book is a free download from Francois Fleuret's website (https://fleuret.org/francois/). There are a bunch of other great resources on his site! Highly recommended to checkout!
The field is ever evolving and static intros formats are thus disadvantaged, likely this will be a fun historical snapshot towards the end of the year. Thanks for keeping it short
The main contribution of the book, I would say, is leaving our the branches of DL which were losing steam for a while now (e.g. bye, RNN's)
This book starts with the basic introduction and understanding of the neural network and gives needed mathematical notions and narration in the field of neural networks . This is very broader for the beginners but this book still gives a good narration or help you to understand the basics of the different neural network models with their applied role in the world
This small book explains quite well what happens in the black box of very classic methods like Nonlinear Algebra, Stochastics, Gradient Descent. At the beginning of the book, it briefly explains how a GPU functions.
A very short discussion on the state of deep learning till about Mid 2023. It is not a practitioner's guide, nor a introduction-to for lay users. Some technical depth is required. I can see it being useful if someone is out of regular touch with the field, and wants a quick brush up.
A great short overview that cites all the papers you can go read if you need to know more about some of the described methods. For a non-CS grad like me, most of the details certainly went over my head, but it‘s a good intro and entry point!
This book is like a dictionary. It helps you recollect the concepts you know and makes you aware of concepts you don’t. It assumes some in-depth knowledge about math and ml as pre-requisites though
Un bon livre pour comprendre les tenants mathématiques des modèles actuels. Bien vulgarisés. Il ne manque que des exemples de codes pour implémenter ça.
The version sold on various platforms such as Amazon, with ISBN 9732346493 is an UNAUTHORIZED COPY of my book. Somebody took the free pdf that I distribute under a non-commercial CC license and sells it. It is an early preprint, sold for 5 times the price of the official version.
The authorized version is available from https://fleuret.org/lbdl as a free phone-formatted pdf, or a $8.50 paperback edition from lulu.com.