M3: Towards efficient mixed machine learning model co-location on constrained edge devices

Author

Luis Angel D. BATHEN
Simeon BABATUNDE
Rhui Dih LEE
Achintya KUNDU
Laura WYNTER, Singapore Management UniversityFollow

Publication Type

Conference Proceeding Article

Publication Date

11-2023

Abstract

This paper explores the effects of machine learning (ML) model co-location on resource-constrained edge devices, and proposes M3, a mixed machine learning model co-location framework that leverages techniques such as network architecture search to generate resource-aware models, as well as quantization to further reduce resource requirements, alongside a runtime that dynamically switches between resource-aware models to serve specific tasks.

Keywords

edge computing, machine learning (ML), object detection, quantization, workload co-location

Discipline

Numerical Analysis and Scientific Computing

Research Areas

Intelligent Systems and Optimization

Publication

Proceedings of the 2023 IEEE Military Communications Conference, MILCOM 2023, Boston, MA, USA, October 30 - November 3

First Page

39

Last Page

44

ISBN

9798350321814

Identifier

10.1109/MILCOM58377.2023.10356363

Publisher

IEEE

City or Country

Piscataway, NJ

Citation

BATHEN, Luis Angel D.; BABATUNDE, Simeon; LEE, Rhui Dih; KUNDU, Achintya; and WYNTER, Laura. M3: Towards efficient mixed machine learning model co-location on constrained edge devices. (2023). Proceedings of the 2023 IEEE Military Communications Conference, MILCOM 2023, Boston, MA, USA, October 30 - November 3. 39-44.
Available at: https://ink.library.smu.edu.sg/sis_research/10344

Additional URL

https://doi.org/10.1109/MILCOM58377.2023.10356363