Get 20% off all career paths from fullstack to AI
Build with Azure OpenAI, Copilot Studio & Agentic Frameworks — Microsoft Certified
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore the challenges and solutions for testing machine learning software components in this insightful PyCon US talk. Delve into the evolving landscape of machine learning integration in everyday applications and the unique testing challenges posed by nondeterministic systems. Learn about the importance of testing in ML-driven software development, including handling variations in model training, managing sensitive data pipelines, and addressing model degradation over time. Discover valuable open-source tools, reusable techniques, and hard-won lessons for building stable and reliable ML-powered applications. Gain practical insights on navigating the complexities of integrating machine learning into software development processes and ensuring the quality and reliability of ML-driven systems.
Syllabus
TALK / Rebecca Bilbro, Daniel Sollis, Mark, Patrick Deziel /PyTesting the Limits of Machine Learning
Taught by
PyCon US