CEMSE Weekly Updates - July 1, 2025 Tue, Jul 1 2025 Newsletter Upcoming Events Stay informed about the upcoming events and the latest news from CEMSE. Modern Privacy-preserving Machine Learning: Rigorous Approach for Data Privacy Zihang Xiang, Ph.D. Student, Computer Science Jul 6, 10:00 - 12:00 B3 L5 R5216 privacy-preserving machine learning Differential privacy Federated learning This dissertation centers around privacy-preserving technologies (differential privacy) in broad machine learning applications. This dissertation focuses on two sides of differential privacy: 1) designing privacy-preserving algorithms, 2) ensuring the falsifiability of privacy claims. Explainability and Efficiency in Spatio-Temporal Models: Applications to Traffic Forecasting Xiaochuan Gou, Ph.D. Student, Computer Science Jul 6, 15:00 - 18:00 B5 L5 R5209 traffic forecasting Graph Neural Networks model interpretability This dissertation addresses key challenges in deep learning-based traffic forecasting, including computational efficiency, model interpretability, and data limitations, despite recent progress in spatio-temporal modeling techniques. Towards Usable and Useful Explainable AI Lijie Hu, Ph.D. Student, Computer Science Jul 7, 17:00 - 19:00 B3 L5 R5220 explainable AI Large Language Models multimodal models This talk presents advancements in Explainable AI, spanning from classical deep learning to large language models, with contributions that enhance both the usability and usefulness of interpretability methods to improve trust, performance, and safety in AI systems. Theory and Implementation of Novel Numerical Methods for Multiphysics Interface Problems Najwa Alshehri, Ph.D. Student, Applied Mathematics and Computational Sciences Jul 8, 15:00 - 17:00 B2/B3 L0 R0215 finite element method numerical analysis fluid-structure interactions mixed finite elements This thesis develops and analyzes Finite Element Methods (FEM) for multiphysics interface problems using the Fictitious Domain with Distributed Lagrange Multiplier (FD-DLM) framework. It introduces new families of stable mixed methods with discontinuous Lagrange multiplier spaces, studies both a priori and a posteriori error estimates, and designs multigrid preconditioners. Theoretical results are supported by numerical experiments. Discovery of Low-Dimensional Generative Models for Complex Dynamical Systems Juan Pablo Muñoz Díaz, Ph.D. Student, Applied Mathematics and Computational Sciences Jul 9, 09:00 - 10:00 B2 L5 R5220; Zoom Meeting 95274807609 generative ai applied mathematics bioscience Dynamical Systems deep learning This thesis presents a data-driven framework for discovering low-dimensional generative models of complex systems by using a library of normal-form equations to identify both observable dynamics and hidden control variables directly from time-series data.