Compositional Optimization for Advanced Machine Learning

Erscheinungsjahr 2026
ISBN: 978-3-032-34357-4
Verlag: Springer

Buch, Englisch, 434 Seiten, Format (B × H): 155 mm x 235 mm

Reihe: Springer Optimization and Its Applications

This book offers a comprehensive exploration of compositional optimization, a cutting-edge paradigm reshaping the landscape of machine learning (ML) and artificial intelligence (AI). As AI systems grow increasingly complex, traditional optimization methods fall short, necessitating innovative approaches to tackle intricate problems. This book bridges this gap, providing a systematic treatment of compositional optimization and its applications in modern AI.

Key concepts such as convex optimization, empirical risk minimization, distributionally robust optimization, stochastic optimization and stochastic compositional optimization are thoroughly examined. The chapters delve into the intricacies of optimization problems that exhibit compositional structures, offering both theoretical insights and practical implementation strategies. Readers will benefit from rigorous analysis, practical tips, and access to Github code repositories, making this book an essential resource for those looking to apply these concepts in real-world scenarios.

Designed for graduate students, applied researchers, and professionals with a foundational understanding of ML, this book serves as both a theoretical guide and a practical toolkit. It is an invaluable resource for anyone interested in the intersection of optimization and machine learning, offering insights that are both deep and actionable.

Zielgruppe

Research

Autoren/Hrsg.

Yang, Tianbao

Fachgebiete

• Mathematik | Informatik Mathematik Numerik und Wissenschaftliches Rechnen Optimierung
• Mathematik | Informatik EDV | Informatik Informatik Künstliche Intelligenz Maschinelles Lernen

Inhaltsverzeichnis

Basics: Convex Optimization.- Introduction: Advanced Machine Learning.- Classic: Stochastic Optimization.- Foundations: Stochastic Compositional Optimization.- Advances: Finite-sum Coupled Compositional Optimization.- Applications: Learning Predictive, Generative and Representation Models.

Über Autor(innen)

Tianbao Yang is a Professor and Stephen Horn ‘79 Engineering Excellence Chair at CSE department of Texas A&M University, where he directs the lab of Optimization for Machine learning and AI (OptMAI Lab). His research interests center around optimization, machine learning and efficient AI with applications in medicine. Before joining TAMU, he was an assistant professor and then tenured Dean’s Excellence associate professor at the Computer Science Department of the University of Iowa from 2014 to 2022. Before that, he worked in Silicon Valley as Machine Learning Researcher for two years at GE Research and NEC Labs. He received the Best Student Paper Award of COLT in 2012, and the NSF Career Award in 2019.    He is recognized for his contributions to optimization in ML/AI. He is the founder of the widely used LibAUC library. His NeurIPS 2013 paper on distributed optimization pioneered the ideas of local updates and model averaging, which later became fundamental to federated learning. He also introduced the empirical X-risk minimization framework along with efficient algorithms for solving it, addressing decades-long open problems in machine learning and forming the foundation of the LibAUC library. He is the author of the book “Compositional Optimization for Advanced Machine Learning”. He is associate editor of multiple journals, including IEEE Transactions on Pattern Analysis and Machine Intelligence and ACM Computing Surveys.