The Optimal Sample Complexity of PAC Learning
arXiv:1507.00473v4167 citations
AI Analysis
This solves a foundational problem in computational learning theory, providing optimal sample bounds for PAC learning.
The paper establishes a new upper bound on the sample complexity for PAC learning in the realizable case, matching known lower bounds up to constant factors, thereby solving a long-standing open problem.
This work establishes a new upper bound on the number of samples sufficient for PAC learning in the realizable case. The bound matches known lower bounds up to numerical constant factors. This solves a long-standing open problem on the sample complexity of PAC learning. The technique and analysis build on a recent breakthrough by Hans Simon.