What you will learn
- An understanding of some of the common challenges in testing trained functionality
- A way of defining functional requirements for trained functionality
- How to qualify trained functionality by using traditional test levels.
Session Details
- Intermediate
- 45 Minutes
- Includes 15mins Q&A
- New Technologies & Challenges
Session Speakers
Martin Karsberg
Senior Tester – Infotiv, Sweden
I have worked within the field of software testing for over two decades, from testing IPv6 implementations in embedded software, to helping IT organizations with challenges in test processes. Since 2018 part of my time has been dedicated to research around the topic of verification and validation of Machine Learning functionality. Over the last couple of years I have tried to give back to the software community by lecturing and speaking at conferences, and hold courses in test methodologies at a local polytechnical college. Proud speaker at EuroSTAR 2019 in Prague 🙂
Elias Sonnsjö
Competence Leader – Invotiv, Sweden
I have been working as a test system developer and engineer for three years, mainly towards virtual testing and verification. My biggest interest is in the field of Machine Learning. It’s the selected focus of my Master’s education and thesis – how to verify the intended functionality of something trained but not coded. In two research projects carried out at Infotiv, I’ve been exploring this, finding argumentative patterns as well as hidden biases and issues concealed in one of today’s “hottest” topics. Sharing is caring, as they say, and I hope to be able to enlighten these issues and discuss them with you.
Stay in the Loop
We want to ensure you never miss important announcements, updates and special offers from EuroSTAR.