2026-09-19T08:11:21Z https://www.ideals.illinois.edu/oai-pmh

oai:www.ideals.illinois.edu:2142/125692 2025-02-10 col_2142_5131 col_2142_10761 com_2142_5130 com_2142_10755 com_2142_234

application/pdf en text Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01 The student, Beomyeol Jeon, accepted the attached license on 2024-07-04 at 12:16. The student, Beomyeol Jeon, submitted this Dissertation for approval on 2024-07-04 at 12:26. This Dissertation was approved for publication on 2024-07-08 at 12:17. DSpace SAF Submission Ingestion Package generated from Vireo submission #20949 on 2025-02-04 at 21:16:21 Machine learning systems in constrained environments Jeon, Beomyeol 2024-07-08 Gupta, Indranil Gupta, Indranil Caesar, Matthew Park, Yongjoo Wang, Chen Machine Learning Systems Machine Learning Constraints Placements Autoscaling Graph Neural Networks Serverless Computing Optimizations Algorithms eng Machine learning (ML) training and inference systems encounter constraints in current computation environments due to increased ML model sizes, the fast-growing popularity of ML/AI, etc. In this thesis, we show how machine learning training and inference systems can be executed successfully and efficiently in constrained computation environments, such as limited-memory GPUs, on-premises clusters, and serverless environments, by using a novel combination of algorithms, optimizations, and well-reasoned system designs. Concretely, we propose (i) a system that enables large ML model training over multiple memory-constrained GPU devices via algorithms and system designs that achieve fast placements with a quality comparable to expert-designed placements, (ii) a system that enables efficient resource sharing among ML inference jobs in fixed-size on-premises clusters by making close-to-optimal autoscaling decisions quickly via several relaxation methods in optimization and prediction, and (iii) a system that enables cost-efficient distributed GNN training on constrained serverless execution environments by auto-tuning configuration via analytic model-based offline optimization and gray-box heuristic-based online optimization. 2024-08 Thesis https://hdl.handle.net/2142/125692 Copyright 2024 Beomyeol Jeon Siebel Computing &DataScience Computer Science University of Illinois at Urbana-Champaign Ph.D. Dissertation