Engineering safer machine learning

Normalized final exposure for µ = 0.1 as a function of , for varying tolerance level α. Larger values of α and achieve lower handicap (which implies faster detection). Credit: IEEE Transactions on Automatic Control (2023). DOI: 10.1109/TAC.2023.3240925

Children first learning to walk may go a bit too fast and fall down, or run into a piece of furniture. However, that cause-and-effect element teaches them invaluable information about how their bodies move through space so that they can avoid falling in the future.

Machines learn in a lot of the same ways that humans do, including learning from their mistakes. However, for many machines—like …

Be the first to comment

Leave a Reply

Your email address will not be published.