Machine Learning and Pure Math - Especially Extremal Combinatorics

Explore the intersection of machine learning and pure mathematics through this 58-minute conference talk focusing on extremal combinatorics. Learn from Jordan Ellenberg's experiences collaborating with both industry and academic partners to leverage contemporary machine learning techniques for mathematical research. Discover how ML methods can construct examples in extremal combinatorics, including large capsets, large subsets of grids without isosceles triangles, and graphs with many edges but no 4-cycles. Understand why extremal combinatorics has proven particularly suitable for machine learning applications compared to other mathematical fields, and examine the unique characteristics that make this area of mathematics especially amenable to ML-driven discovery and problem-solving approaches.