Summary

In this code pattern, learn how to utilize AutoAI to immediately produce a Jupyter Note pad which contains Python code of a device learning model. Then, explore, modify, and retrain the design pipeline using Python before deploying the model in IBM Watson ® Artificial intelligence using Watson Machine Learning APIs.

Description

AutoAI is a graphical tool readily available within IBM Watson Studio that analyzes your data set, produces numerous design pipelines, and ranks them based upon the metric chosen for the issue. This code pattern reveals extended functions of AutoAI. More standard AutoAI expedition for the exact same information set is covered in the Generate device learning model pipelines to choose the very best model for your problem tutorial.

When you have finished this code pattern, you understand how to:

  • Run an AutoAI experiment
  • Generate and conserve a Python note pad
  • Execute the note pad and analyze results
  • Make modifications and retrain the model using Watson Artificial intelligence SDKs
  • Deploy the model utilizing Watson Artificial intelligence from within the notebook

Flow

flow

  1. The user sends an AutoAI experiment using default settings.
  2. Numerous pipeline models are created. A pipeline model of choice from the leaderboard is conserved as a Jupyter Notebook.
  3. The Jupyter Note pad is executed, and a modified pipeline design is created within the note pad.
  4. The pipeline model is released in Watson Machine Learning using Watson Machine Learning APIs.

Guidelines

Get in-depth instructions in the readme file. These instructions discuss how to:

  1. Run an AutoAI experiment.
  2. Save the AutoAI-generated notebook.
  3. Load and perform the note pad.
  4. Release and score as a web service using a Watson Machine Learning instance.