An improved uniform convergence bound with fat-shattering dimension

Roberto Colomboni, Emmanuel Esposito, Andrea Paudice

Abstract: Highlights•This paper improves state-of-the-art bounds on the sample complexity for uniform convergence of real-valued functions.•Uniform convergence is a fundamental tool in the design and analysis of machine learning algorithms.•Our work builds upon chaining and packing number arguments to provide optimal sample complexity bounds.