DataEval Tutorials¶

The tutorials on this page aim to teach important concepts one can come across as a machine learning testing and evaluation engineer through easy-to-follow examples that are handcrafted to give you the best experience possible.

The tutorials are split into sections based on which stage they fall under in the machine learning life cycle. Currently, there are tutorials for the following stages:

Data engineering
Monitoring

More tutorials are always in development, but you can suggest specific tutorials by requesting a topic.

To view the tutorial directly in the browser, click the title.

To run the tutorial interactively in Google Colab, click the icon.

Data engineering¶

Introduction to Data Cleaning	Learn about the impacts of unstructured, raw data and how to transform it into a reliable, robust dataset.
Assess an unlabeled data space	Learn how to fix and prevent gaps in data to develop more reliable and robust models.
Identify bias and correlations	Learn how correlations in your data and metadata can affect model performance and what can be done to remove that bias.

Monitoring¶

Monitor shifts in operational data	Learn how to analyze incoming data against training data to ensure deployed models stay reliable and robust.