ifanfan / AutoML-AIConf-2019-BJ

AutoML tutorial for AI Conferencen Beijing 2019

Artificial Intelligence Conference 2019 Beijing

Tutorial: Democratizing and accelerating AI through automated machine learning

Click here to see session details.

0. Prerequisite

To make most of the tutorial session, please follow the following steps and set up the environment:

1. Open Azure Portal
2. Open the provided link and enter the activation code, then sign up.
3. Click the workspace as shown below

If this workspace is not shown, click All resources on the left then click the only Machine Learning service workspace as shown below

4. Click Download config.json and save config.json as shown below

5. Open Azure Notebooks and click My Projects as shown below

6. Click Upload GitHub Repo as shown below

7. Follow the instruction below to upload this repo to your Azure Notebooks and click Import (The repo link is henry-zeng/AutoML-AIConf-2019-BJ)

8. Upload config.json to the repo you imported just now as shown below

9. Click 0_setup.ipynb, and run the whole project as shown below

If asked to select a Kernal, please choose "Python 3.6"

👏 Congratulations! You are all set! 👏

1. Outline

In this session, we are going to run a few notebooks to learn:

• What are the strengths of AutoML
  
  and how to utilize AutoML to do:
• Training and hyperparameter tune with TensorFlow
• Explain classification model and visualize the explanation
• Classification with ONNX
• Classification with Deployment

2. Feedback

Thank you for joining us in today's tutorial session! We hope you enjoyed it! Welcome to leave any feedback, comments and thoughts here.