Approaches for evaluating functions over distributed data streams are increasingly important as data sources become more geographically distributed. However, existing methodologies are limited to small classes of functions, requiring non-trivial effort and substantial mathematical sophistication to tailor them to new functions.
In this talk I will present AutoMon, the first general solution to this problem. AutoMon, first presented in SIGMOD 2022, enables automatic, communication-efficient distributed monitoring of arbitrary functions. Given source code that computes a function from centralized data, the AutoMon algorithm approximates the function over the aggregate of distributed data streams, without centralizing data updates.
Our evaluation shows that AutoMon sends the same number or fewer messages as state-of-the-art techniques when monitoring specific functions for which a distributed, hand-crafted solution is known.
AutoMon, however, is a lot more powerful. It automatically generates a communication-efficient distributed monitoring solution for arbitrary functions, e.g., monitoring deep neural networks inference tasks for which no non-trivial solution is known.
Moshe (Mickey) Gabel is an assistant professor in the Department of Electrical Engineering and Computer Science at the Lassonde School of Engineering, York University. Before joining York, he spent four years as a limited-term assistant professor in the Department of Computer Science at the University of Toronto. Moshe earned his PhD in Computer Science from the Technion - Israel Institute of Technology, where he also got his MSc and BSc.
Moshe's research lies in the intersection of distributed algorithms, systems, and machine learning. His current research interest is edge computing, specifically making geo-distributed data analysis more practical and accessible to typical software developers. He has also worked extensively on machine learning applications in pervasive health monitoring and in computer systems. Moshe's work appeared in top venues for systems and data science, including SIGMOD, NSDI, SIGKDD, VLDB, and ICML. He has served on multiple program and organizing committees, most recently as program chair for SYSTOR '22 and PC member of NeurIPS '22 and ICML '22.