Topic: "Unveiling Collective Intelligence: Navigating Representative Learning for Federated Insights"
Presenter: Keren Li, Ph.D.
Assistant Professor of Mathematics
College of Arts and Sciences
University of Alabama at Birmingham
Abstract: In the ever-evolving landscape of modern data exploration, the pursuit of collective intelligence has ushered in innovative paradigms that transcend conventional data analysis methods. Distributed learning, a fusion of big data and machine learning, orchestrates learning across nodes distinguished by their unique characteristics, including massive distribution, non-iid data, data imbalance, and constraints imposed by privacy and limited bandwidth. These multifaceted challenges disrupt the conventional norms of machine learning, necessitating innovative solutions that can harmonize disparate data perspectives while addressing critical concerns of privacy and resource constraints.
Representative Learning plays a pivotal role by crafting pseudo data points, known as representatives, which encapsulate the essential features of local data nodes. The representative sets are channeled to train regular models at the central unit, enabling a symphonic convergence of insights.
Representative Learning, with its ingenious architecture, serves as conduits that transcend the barriers posed by data privacy and scarce communication resources, sidestepping concerns that often hinder analysis. Moreover, distributed learning thrives in the heterogeneity between node distributions, particularly in settings characterized by vast node numbers. Smaller nodes inherently yield more reasonable representatives, while the inherent dissimilarity between nodes safeguards low variance of estimation from the representative learning approach.
Friday, September 29 at 10:00am to 11:00amVirtual Event