Dr Alessandro De Palma

简介

A large body of his work leverages network over-approximations to formally prove (neural network verification) and enforce (robust training) properties pertaining to the safety and reliability of neural networks.