FactSheets for AI services are like nutrition labels for food. Concerns about safety, transparency, and bias in AI are widespread. Part of the problem is a lack of standard practices to document how an AI service was created, tested, deployed and evaluated; how it should operate; and how it should (and should not) be used. The AI FactSheets concept provides a framework to document machine learning models and AI services. Learn how to document models using FactSheets, taking two models as examples. By the end of the session system architects, application developers as well as data scientists will gain an understanding for how to describe, compare, assess and select models for a particular use case or application.
🎓What you will learn
In this hands-on workshop developers will learn:
- What are factsheets?
- Why it is important?
- Methodology for creating factsheets
- AI Lifecycle Governance
- Resources to know more about the factsheet
👩💻 Who should attend
Developers, Data Scientist or anyone who is interested in learning about factsheet and AI Governance
Saishruthi Swaminathan, Data Scientist and Developer Advocate, IBM
Upkar Lidder, Data Scientist and Developer Advocate, IBM