Computable PAC Learning - Part 1

Explore a detailed seminar talk from the historic Kolmogorov seminar series on computational and descriptional complexity, where Tomasz Steifer from the Pontifical Catholic University of Chile presents the first part of his research on computable PAC (Probably Approximately Correct) learning. Delivered as part of the long-running seminar series established by Kolmogorov around 1979, this 85-minute presentation delves into the theoretical foundations and computational aspects of machine learning through the lens of PAC learning frameworks.