Por Alex Usvyatsov (Universidade de Lisboa, CMAF-CIO).

Abstract: I will discuss the seminal theorem of Vapnik and Chervonenkis on the uniform convergence of relative frequencies of events to their probabilities (a uniform version of the weak law of large numbers). This result is vastly regarded as the foundation of machine learning. However, it has interesting implications in many other areas of mathematics. I will discuss some connections between VC-theory and logic. As promised in one of my previous talks, I will also say something about the proof of the theorem.

This talk is completely independent of the previous ones.