TUTORIAL-2
TITLE: Federated Learning – Introductory Tutorial
ABSTRACT: Federated Learning (FL) is a modern paradigm in the field of machine learning, enabling collaborative model training across multiple decentralized agents while preserving data privacy. FL is particularly important when data must be stored locally (e.g., for privacy) and there is a need to collaboratively train a common model across connected learning agents, achieving superior performance compared to what each agent could achieve using only its own limited local data. FL can find applications in numerous areas, including healthcare, finance, and situational awareness in military and industrial operations. This tutorial aims to provide an overview of FL, from its foundational principles to state-of-the-art techniques. Attendees will gain insights into important challenges in FL, including data distribution, communication efficiency, and privacy preservation. The tutorial will delve into advanced topics such as algorithms for FL training/optimization, handling data heterogeneities, and enhancing efficiency. Additionally, practical implementation aspects, real-world applications, and future directions of FL research will be discussed.
INSTRUCTOR: Prof. Panagiotis (Panos) Markopoulos, PhD, Margie and Bill Klesse Endowed Associate Professor, with the Departments of Electrical and Computer Engineering and Computer Science at The University of Texas at San Antonio (UTSA). He is a core faculty member of the UTSA School of Data Science and MATRIX: The UTSA AI Consortium for Human Well-Being. At UTSA, Dr. Markopoulos has founded and directs the Machine Learning Optimization Research Lab (MELOS) and the Multimodal Sensing and Signal Processing Teaching Lab (MSSP). Before joining UTSA, Dr. Markopoulos was a tenured Associate Professor at the Rochester Institute of Technology (RIT). He was also a Visiting Research Faculty at the U.S. Air Force Research Laboratory (AFRL) in Rome, NY, during the summers of 2018, 2020, and 2021. His expertise spans machine learning, data analysis, and adaptive/statistical signal processing, with a focus on machine learning from challenging data, such as corrupted, streaming, imbalanced, missing/sparse, multimodal, and privacy-sensitive data. His research mission is to advance efficient, explainable, and trustworthy artificial intelligence for the benefit of our society and nation. In 2019, he received the Young Investigator Program (YIP) Award from the U.S. Air Force Research Office (AFOSR), to conduct research on dynamic and robust subspace analysis of tensor data.