Machine learning (ML) models are highly flexible but face fundamental limits when learning from finite data. We show that a universal lower bound—the Limits-to-Learning Gap (LLG)—captures the unavoidable gap between the true (population) benchmark and linear ML models.
Memories of the LQG 2025/09/03 – 2025/09/06 – Autumn Seminar – at Cambridge – to immortalise the speakers; inspire members who did not attended; and reawaken fond memories among those who did.
AI : What is real, and what is hype? What capabilities? What can’t they do? Explore the current state of the art in AI applications
Highlight the utility of deep reinforcement learning in finance and invent “economic distillation” tools for interpreting AI and big data models