DeepRead
Cover of Deep Learning from Scratch

Deep Learning from Scratch

Seth Weidman · 2019

Machine learning, statistics and probability › Deep learning and AI

#4,246 most recommended

About the book

A practical introduction to neural networks and deep learning that builds the core machinery in Python with NumPy rather than relying on a framework. It starts from simple models such as linear regression and works up to a small library of layers, activations, loss functions and optimizers, with later chapters covering convolutional and recurrent networks. Deep learning becomes understandable once the mechanics are implemented directly: a network is a composition of simple operations, and training it means computing how the loss changes with each parameter through the chain rule and backpropagation, then adjusting parameters by gradient descent. The book claims that a modest amount of code is enough to reproduce how modern frameworks work, so that a reader who has built these pieces can reason about the behaviour of larger models and libraries.

Written by Claude from what it knows of this book.

What Hacker News says 44% recommend · 11% critical

02.5520192020202220242022 peak · 5
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.

  • 12 want to read
  • 1 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
2019
ISBN
9781492041412
Editions known
2 at Open Library
Open Library
OL21334210W

Explore in the mosaic

Read next: its neighbours on the mosaic