Speaker: Chi-Huan Tung
Title: Probabilistic Inference for Enhanced Scattering Data Analysis
This presentation introduces a unified probabilistic inference framework for extracting physically meaningful quantities from experimental measurements. Rooted in linear algebra, the method systematically propagates uncertainties and reconstructs latent variables from sub-optimal data using matrix-based formulations. Special attention is given to scenarios where the target quantity is not directly observable, but is instead linked to the data through known forward or inverse linear operators. To demonstrate the generality and effectiveness of this approach, we present case studies drawn from real-world scattering experiments, including Small-Angle Neutron Scattering (SANS), Neutron Spin Echo (NSE), and X-ray Photon Correlation Spectroscopy (XPCS).
Host: Koichi Mayumi
