About this Event
As we dive into the realm of RL, where autonomous agents learn to make decisions through trial and error, ensuring their safety and reliability becomes paramount. Neural networks serve as the backbone of these agents, learning complex mappings from observations to actions.
Formal verification offers a systematic approach to assessing the behavior of these neural networks. By subjecting them to rigorous mathematical analysis, we can guarantee properties such as safety, stability, and fairness. This becomes especially crucial in high-stakes environments like autonomous vehicles, medical diagnosis systems, and robotics.
Throughout this session, we'll explore how formal verification techniques provide a safety net against unexpected behaviors or vulnerabilities in RL systems.
Event Venue & Nearby Stays
Ca' Vignal 3, Strada le Grazie, Verona, VR, Italy
EUR 0.00
