按年被提到
20162017 · 28 次2026
开源书›机器学习与 AI·No.14

Deep Learning

深度学习
Ian Goodfellow, Yoshua Bengio, Aaron Courville

Goodfellow 等人合写的深度学习教科书,从数学基础讲到现代网络与研究前沿

免费阅读 ↗
被推荐 · 93 人首次 2016-02-24 · 最近 2025-08-14官方免费在线(版权保留,作者或出版社免费放出)
93不同的人
57活跃月份
45有推荐的月份
95讨论串

由 Ian Goodfellow、Yoshua Bengio、Aaron Courville 合著,先用几章复习线性代数、概率和数值计算,再讲前馈网络、正则化、优化、卷积网络、循环网络,最后进入生成模型等研究性话题。写法偏教科书和综述,重理论与直觉,没有配套的代码练习。和偏动手的入门课程相比,它更系统、更偏数学,适合当作参考和打基础用。

HN 为什么推荐它

从 2016 年起,HN 上的人常把它当作“深度学习圣经”推荐:免费、全面,数学复习部分也比较完整,适合在入门课之后深入读;也有人说自己没读完,只读过的章节很好。

难度 4 / 5约 80 小时
适合谁

想系统打牢深度学习理论基础的人

读之前最好知道
1线性代数
2概率论
3微积分
4Python 编程

怎么读

由 Claude 根据这本书和 HN 上的说法整理,可能有错
  1. 1数学基础不牢先读开头的数学复习部分,已经熟悉可以跳过
  2. 2先读前馈网络、正则化、优化,再读卷积和循环网络
  3. 3配合一门动手课程,用代码实现正在读的内容
  4. 4后面的研究性章节可以按兴趣选读,不必从头读到尾

HN 书评

推荐在前,同类按时间 · 每本最多保留 20 条,统计用全部提及 · 作者本人贴的不算
强烈推荐j_juggernaut · 2016-09

“I highly recommend it if you want to get deeper intuitions on the principles of CNN, LSTM/RNN, Restricted Boltzmann Machines etc.”

强烈推荐,能深入理解 CNN、RNN、RBM 等原理
Neural Network Architectures原评论 ↗
推荐bronz · 2016-03

“For people who want to learn more about machine learning, check out this free textbook written by Yoshua Bengio.”

推荐 Bengio 参与写的免费机器学习教材
Enrollment Is Surging in Machine Learning Classes原评论 ↗
推荐hiddencost · 2016-03

“Deep learning seems popular these days:)”

拿它作为深度学习的教材资源
Enrollment Is Surging in Machine Learning Classes原评论 ↗
推荐curuinor · 2016-05

“Bengio's Deep Learning book is the most current by far.”

称是关于神经网络最新的一本书
Ask HN: Best book on machine learning?原评论 ↗
推荐max_ · 2016-06

“- 5. Finnish off. Deep Learning”

把它列为学习路径的最后一步
Ask HN: From programming to AI, how?原评论 ↗
推荐anantzoid · 2016-06

“If you want to read one main resource... the Goodfellow, Bengio, Courville book (available for free from Deep Learning) is an extremely comprehensive survey of the field.”

推荐作为唯一主读物,称是全面的领域综述
How to Start Learning Deep Learning原评论 ↗
看全部 20 条 →
推荐Dzugaru · 2016-06

“Following that Ian Goodfellow book Deep Learning is very good.”

推荐在入门 ML 课程之后读 Goodfellow 这本
How to Start Learning Deep Learning原评论 ↗
推荐Uptrenda · 2016-07

“Here's a good book on convolutional neural networks: Deep Learning It's suppose to be like the bible on deep learning at the moment”

推荐为 CNN 的好书,称是目前深度学习的圣经
A Course in Machine Learning原评论 ↗
推荐kmike84 · 2016-08

“For deep learning check an awesome Deep Learning and try to implement what you're reading about”

推荐,建议边读边动手实现
Facebook AI Research Team Open Source DeepMask and SharpMask原评论 ↗
推荐brhsiao · 2016-10

“The chapters I read from that were very good, but I didn't finish the book so hesitated to recommend it.”

称读过的章节很好,但没读完所以不敢推荐
Introduction to Convolutional Neural Nets原评论 ↗
推荐momeara · 2016-10

“Deep Learning is a wonderful resource to learn more about why and when to use different architectures.”

推荐用来理解不同架构为何用、何时用
Keras-based molecular autoencoder原评论 ↗
推荐imakecomments · 2016-11

“See: Deep Learning for better treatment of the math review. It seems more thorough and makes less assumptions about the math background of the reader.”

推荐其数学复习部分,更详尽且少假设读者背景
Tensorflow and Deep Learning, Without a PhD, Martin Gorner, Google [video]原评论 ↗
推荐adamnemecek · 2016-11

“* Deep Learning (Deep Learning or (链接))”

把它列为中级的 ML 学习资源
Machine Learning for Software Engineers原评论 ↗
推荐throwaway001231 · 2016-11

“If you are serious about deep learning I would recommend this book instead: Deep Learning”

推荐给认真学深度学习的人,全书可在线看
Machine Learning for Software Engineers原评论 ↗
推荐theCricketer · 2016-12

“3. More theoretical background: Deep Learning”

列为更偏理论背景的读物
Learning Machine Learning: A beginner's journey原评论 ↗
推荐inlineint · 2017-01

“I like the lectures by Nando de Freitas ((链接)) and a book by Ian Godfellow.”

推荐 Goodfellow 的书,与 de Freitas 的讲座并列
Ask HN: What languages/technologies do you plan to learn in 2017?原评论 ↗
推荐TheAlchemist · 2017-01

“There is also this one (not really a tutorial though):”

补充推荐,称它不算教程
Stanford Unsupervised Deep Learning Tutorial (2014)原评论 ↗
推荐gregatragenet3 · 2017-02

“If the video above leaves you wanting more, Deep Learning does a good job on drilling into more specifics for the various techniques used.”

推荐它深入讲解各种 deep learning 技术的细节
TensorFlow 1.0 Released原评论 ↗
推荐dongslol · 2017-03

“Get your feet wet:”

推荐作为上手神经网络的早期资源
Ask HN: Best books on AI?原评论 ↗
推荐laughingman2 · 2017-03

“2. Read Deep Learning to gain the relevant math behind it.”

推荐读它补足背后的数学
Opening a new chapter of my work in AI原评论 ↗