HEP Division Seminar Speaker Tanvi Wamokar (SLAC)

Thursday Oct 22, 2026, 11:00 AM → 12:00 PM America/Chicago

F-108 (Bldg 362)

Title: Machine Learning for Precision Measurements in High-Energy Physics

Abstract: Machine learning is transforming how we perform measurements in high-energy physics, enabling unbinned analyses and more powerful inference. In this talk, I will present two threads of my work in this space. The first focuses on unbinned unfolding and likelihood ratio estimation methods, including contributions to the H1 DIS unbinned measurement and open-source tools developed to broaden community access to these techniques. The second explores the role of physics priors in machine learning: when should physical symmetries and constraints be encoded explicitly into network architectures, versus learned implicitly from data? I will present studies on this tradeoff in the context of likelihood ratio estimation and anomaly detection, along with ensemble strategies to control bias and variance in neural estimators.