Today,, this site recommends a deep learning cheat sheet: the deep learning handout by Professor Li Hongyi of National Taiwan University, which is the easiest introduction to deep learning I have ever seen. The 300-page handout can systematically and easily explain the basic principles of deep learning, just as vivid as the machine learning cheat sheet.

Note: The machine learning cheat sheet of this website was published in the past.

Make machine learning concepts as handy as a cheat sheet for memorizing TOEFL words! Machine learning all kinds of memorized concepts in minutes!

This paper gives a brief introduction to the main content of the cheat sheet, and makes page number annotation and title translation for the main part:

1. Outline of lectures:

Part I: Introduction to deep learning \

1.1 Introduction to Deep Learning (P5)

1.1.1 Three Steps of Deep Learning (P10)

1) Define a series of functions (P11)

1.1.2 Advantages of Functions (P26)

1) Training data (P27)

2) Learning Objectives (P28)

3) Loss function (P29)

1.1.3 Selecting the Best Function (P32)

1) Gradient descent (P33)

2) Derivation of back propagation (P44)

1.2 Why depth is used (P47)

1.2.1 More Parameters for Better Performance (P47)

1.2.2 Any function can be implemented through a single hidden layer (P44)

1.2.3 Deep learning: Modularity? Less data required (P52)

Part II: Some Suggestions on training deep neural networks \

2.1 Appropriate Loss Function (P69)

2.1.1 Square error and cross entropy (P74)

2.2 Mini Batch (P74)

2.2.1 Better Performance (P83)

2.3 Activation Function (P86)

2.3.1 RELU (P92)

2.3.2 Maxout (P98)

2.4 Adjustment of learning Rate (P98)

2.4.1 Adagrad (P104)

2.5 Momentum (P108)

2.5.1 Adam (P112)

2.6 Overfitting solved (P115)

2.6.1 More Training Data (P116)

2.7 Early Stop (P119)

2.8 Weight Attenuation (P121)

2.9 Dropout (P126)

2.10 Network Architecture (P138)

Part 3: Various Neural networks \

3.1 CNN (P149)

3.1.1 Convolution (P158)

3.1.2 Pooling (P165)

3.1.3 Tiling (P170)

3.2 RNN (P192)

3.3 LSTM (P196)

3.4 GRU (P211)

Part 4: The Next Wave \

4.1 Supervised Learning (P226)

4.1.1 Super Deep Network (P226)

4.1.2 Attention Model (P235)

4.2 Enhanced Learning (P252)

4.3 Unsupervised Learning (P264)

2. Screenshot of lecture notes:

3. Conclusion:

One day to understand deep learning, although it is a bit exaggerated, this handout does cover the basic concept of deep learning. After reading the handout, you will have an overall understanding of deep learning, which will be of great help to future learning. Although the handout is not the latest (this version is from June 2017), but for beginners, there will be twice the result with half the effort, this site highly recommended.

Note: For the convenience of readers, this site will convert the PDF version of the handout into PPT format, convenient for everyone to use (some formula is wrong, please refer to the PDF version)

Please reply to “Li Hongyi” for complete download of lecture notes

You can also download it directly with Baidu Cloud:

Link: pan.baidu.com/s/1t0EpHwx4…

Extraction code: H74O

Please follow and share ↓↓↓

Machine learning beginners \

QQ group: 654173748

Past wonderful review \

  • Machine learning for Beginners

  • Printable version of Machine learning and Deep learning course notes \

  • Ng personal notes on machine learning course online edition \

  • DeepLearning. Ai DeepLearning course notes online edition

  • Introduction to Deep Learning – Python Deep Learning, annotated version of the original code in Chinese and ebook

  • Zotero paper management tool

  • The mathematical foundations of machine learning