DeepRead
Cover of Dive into Deep Learning

About the book

An open-source textbook on deep learning that covers the mathematical preliminaries (linear algebra, calculus, probability, optimization) and then neural network methods, from multilayer perceptrons through convolutional networks, recurrent networks, attention, and Transformers. Its scope extends to applications in computer vision, natural language processing, recommender systems, and generative models, along with the computational performance issues involved in training them. The book's central claim is that deep learning is best understood by combining mathematical formulation, runnable code, and small experiments in one document, so that each concept is implemented and tested rather than only described. It treats models as compositions of layers, losses, and optimizers that a reader can build from scratch and then adapt, and it presents modern techniques such as attention alongside the older foundations they build on.

Written by Claude from what it knows of this book.

What Hacker News says 86% recommend · 3% critical

0510201920202022202420252023 peak · 10
This bookAverage book in Deep learning and AI

In the commenters’ own words

What Open Library readers say

No Open Library reader has rated it yet.

  • 3 want to read
  • 0 reading
  • 0 have read

Open Library’s page for the book

What critics say, via Wikipedia

No Wikipedia article about this book yet, so no critics to quote.

Reddit communities — coming later

The threads it came up in

Details

First published
2023
ISBN
9781009389433
Editions known
4 at Open Library
Open Library
OL35722907W

Science

Explore in the mosaic

Read next: its neighbours on the mosaic