Top 10 Countdown: Session #7 — Ensuring Machine Learning Generalization in Avionics Using Formal Methods
Featuring: Dr. Arthur Clavière, Collins Aerospace
How can we be confident that a machine learning model will behave safely on data it’s never seen—especially in avionics? In this session, Dr. Arthur Clavière introduces a formal methods approach to verifying maching learning generalization. The talk highlights how formal verification can be apploied toneural networks in safety-critical avionics systems.
💬 Discussion question:
Where do you see formal verification having the biggest impact on deploying ML in safety‑critical systems—and what challenges still stand in the way?
Join the conversation below 👇
1 Comment
Time Descending
I think formal verification could have its biggest impact in defining clear safety boundaries for ML systems, especially in areas like avionics where worst case behavior matters more than average accuracy. The hard part, it seems, is scaling those verification methods to modern deep networks without making them too simplified to be useful. I’d be really interested in how you see that balance playing out in practice.
Sign in to participate
