Slides from MXNet tutorials
To learn more about Apache (incubator) MXNet Gluon, try our interactive book on Deep Learning and MXNet, taught entirely through runnable Jupyter notebooks. The project is being developed in the open and is available for free https://github.com/zackchase/mxnet-the-straight-dope/
ICML2017 has 104 slides, for a 2-hour tutorial, covering these topics:
- Deep Learning 101
- NDArray
- Multilayer Perceptrons
- Convolutional Neural Networks
- Generative Adversarial Networks
- LSTMs
- TreeLSTMs
- Distributed Optimization
- Parallel Training
From https://2017.icml.cc/Conferences/2017/Schedule?showEvent=894 (Aug 6th, 2017)
Video available here
KDD2017 is 147 slides, for a 3-part, 5 hour tutorial.
1. Getting Started
1.1. Installation
1.2. Deep Learning 101
1.3. NDArray
2. Neural Networks in MXNet
2.1. Linear Models & MLPs
2.2. Loss Functions
2.3. Convolutional Neural Networks
2.4. Optimization & Capacity Control
3. Advanced Topics
3.1. Generative Adversarial Networks
3.2. LSTMs
3.3. TreeLSTMs
4. Scaling Learning
4.1. Infrastructure on AWS
4.2. Parallel Optimization Algorithms
4.3. Parallel and Distributed Training
From http://www.kdd.org/kdd2017/ (Tue, Aug 15 2017)