How to Test Artificial Intelligence and Machine Learning Solutions?
summary
Practical guidance for testing machine learning and AI services: names types of problems that differ from regular software testing and describes approaches based on real delivery experience. Covers how to test deployed ML/AI components, what specific issues to look for, and gives a starting point and methods to validate these systems rather than speculative claims about ML replacing testers.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 2:42ML myths vs reality
- 3:23AI testing stages
- 7:21Titanic data model
- 10:53Feature engineering prep
- 13:46Logistic regression training
- 16:46Model evaluation
- 21:30Feature tuning
- 25:08Model optimization selection
- 26:52Model deployment
- 28:16Prediction microservice
- 30:00Unit testing
- 31:24API integration testing
- 33:31Research vs production
- 35:39Reproducibility logging
- 37:08Model decay monitoring
- 38:52Further learning
- 43:16Q&A
related talks
01h 04mWhat Testers Need from Machine Learning and Artificial Intelligence
51mHow to effectively test your AI applications?
07m“AI ML in Test Automation, Are you ready?”
32mQuality engineering in the AI age: From testing AI to leveraging GenAI
51mDebugging AI
30mAI in QA. Rise of the Machines?
44mApplying AI Capabilities in Scenario Testing
29mTesting and Assuring AI-Based Systems