https://github.com/shujian2015/dl-book-annotated-keras
Annotated Keras - in process
https://github.com/shujian2015/dl-book-annotated-keras
Last synced: 3 months ago
JSON representation
Annotated Keras - in process
- Host: GitHub
- URL: https://github.com/shujian2015/dl-book-annotated-keras
- Owner: Shujian2015
- License: mit
- Created: 2018-12-12T08:31:32.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2018-12-17T03:08:02.000Z (over 6 years ago)
- Last Synced: 2025-02-15T12:50:01.757Z (3 months ago)
- Homepage:
- Size: 38.1 KB
- Stars: 4
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## Annotated Keras
Trying to connect the deep learning book by Goodfellow and the Keras source code to have better understanding of neural network. Inspired by the awesome [annotated transformer post](http://nlp.seas.harvard.edu/2018/04/03/attention.html) from HarvardNLP. This is not a foundamental tutorial. Reader may need to have basic understanding of neural network and Keras. This is for those who want to know how Keras implements neural network and have better understanding of neural network.
### Main Sources
* Deep Learning by Goodfellow et al.: http://www.deeplearningbook.org/
* Keras (2.2.4): https://github.com/keras-team/keras
* Keras doc: https://keras.io/
* Tensorflow: https://github.com/tensorflow/tensorflow
### Table of Contents:
* 6 Deep Feedforward Networks
* 7 Regularization for Deep Learning
* 8 Optimization for Training Deep Models
* 8.1 How Learning Differs from Pure Optimization
* 8.2 Challenges in Neural Network Optimization
* [8.3 Basic Algorithms](ch08/8.3.MD)
* [8.4 Parameter Initialization Strategies](ch08/8.4.MD)
* [8.5 Algorithms with Adaptive Learning Rates](ch08/8.5.MD)
* 8.6 Approximate Second-Order Methods ([issue](https://github.com/keras-team/keras/issues/460))
* [8.7 Optimization Strategies and Meta-Algorithms](ch08/8.7.MD)
* 9 Convolutional Networks
* 10 Sequence Modeling: Recurrent and Recursive Nets