Sidebar[Skip]

Virtual Special Session 14:
Handling Resource constraints for/using ML (CONSTRAINT 2023)

The development of machine learning (ML) technology has increased human efficiency, accessibility, and safety. However, these technologies are often constrained by limitations in computing power, energy, data, and security risks in turn leading to a lack of equitable distribution of technology that is essential for promoting innovation and improving the lives of people. As such “advancing technology for humanity as a whole” requires addressing these resource constraints using a multi-disciplinary approach that involves advances in computing, energy, connectivity, data engineering, and environmental sustainability. As machine learning systems become more prevalent, it is important to develop strategies that enable these systems to operate efficiently [1-3] and sustainably in a wide range of environments and applications [4]. The proposed special session aims to provide a common ground to showcase recent advancements in developing ML systems under various constraints, build collaborations across disciplines, share benchmarks, algorithms, and methods for resource-constrained ML systems, and inspire research in domains where ML is used to address those research constraints.

The broader objective of CONSTRAINT-2023 will be: to investigate challenges and resource constraints that prevent development and access to ML Applications; to promote research in developing equitable ML systems and applications; to provide opportunities for researchers from the area of the “Developing ML Technology Under Resource Constraints” community from around the world to collaborate with other researchers; and to foster the development of a broader multi-disciplinary framework to adopt ML systems in daily life.

Scope and Topics:

CONSTRAINT-2023 welcomes theoretical and practical paper submissions on various scopes to contribute in developing ML systems under a variety of resource constraints and to address resource constraints. We will particularly encourage studies that address either practical applications or improve upon resource constraints for a variety of ML systems in the field.

We invite submissions on topics that include, but are not limited to, the following:

• Creating new resources such as data, hardware, and protocols for ML systems, Algorithms and Applications
• Optimizing data science systems, embedded platforms, and test beds for ML systems, Algorithms and Applications
• Data privacy, hardware privacy, and new ML system and Algorithms design
• Algorithms for Urban computing and ML analytics under resource constraints.
• Energy-efficient computing and inferencing for ML systems, Algorithms and Applications
• Optimized VLSI and architecture design for ML and data science applications
• Algorithms and systems for increasing database efficiency using Machine Learning
• Optimizing Machine Learning Algorithms for environmental sustainability and Green Machine Learning.
• Cheaper surrogate ML systems, Algorithms and Applications
• Tools and methods for “green Machine Learning systems” hardware-software system design and evaluation
• Frameworks and methods to improve equity of Machine Learning systems especially under constraint of data and infrastructure.
• Machine Learning Algorithms catering to applications in resource constraint third world applications.
• Empirical study of resource constraints in areas of healthcare, supply chain, enterprise mobility solution, mobile systems, edge computing, education, smart campus, smart city and buildings, manufacturing, energy, demand forecasting, finance, retail, social computation, crowd sensing, wireless communication and networking, smart mobility, cyber security, environmental policy, climate change, and control, internet of personalized things, etc.
• Applications of AI and ML under resource constraints.

Chair: Manikandan Ravikiran, Soumen Biswas, Sathyanarayanan N. Aakur, Arunkumar Bagavathi

Bio: Manikandan Ravikiran is a Research Scientist working at Artificial Intelligence Research Group at R&D Centre, Hitachi India to better understand how to improve software products of ADAS, Healthcare and Safety using NLP and Computer Vision under resource constraint settings. His research focus includes deep compression, test time generalization and less data learning for resource constrained AI problems. Dr. Soumen Biswas is Senior Researcher in the Artificial Intelligence Research Group at Hitachi R&D India. His research area includes machine learning and deep learning for image classification and segmentation, biomedical image processing, optimization of image energy function, PDE based image segmentation, convolution neural network in image segmentation and other areas in computer vision and artificial intelligence. Dr. Biswas has published various articles in journals and presented papers in conferences of repute. Dr. Sathyanarayanan Aakur is an Assistant Professor in the Department of Computer Science at Oklahoma State University. His research interests lie at the intersection of artificial intelligence and psychology, particularly in the use of common-sense reasoning in multimodal understanding. Much of current work focus include self- supervised predictive learning, common-sense reasoning to ground perception and prior knowledge, generative modelling for building knowledge from the ground-up. He is a recipient of NSF CAREER Award in 2022. His research has been published in venues such as IEEE CVPR, AAAI, IEEE WACV, Quarterly of Applied Mathematics, etc. Dr. Arunkumar Bagavathi is an Assistant Professor in the Department of Computer Science at the Oklahoma State University. He received his doctoral degree from UNC Charlotte in 2019. His research interests include data mining, graph mining, computational social science, and applied machine learning. His research works have been published in venues like IEEE Big Data, Complex Networks, IEEE/ACM ASONAM, IEEE ICMLA, and JMIR.

Technical Committee (tentative)

• Ratnavel Rajalakshmi, Vellore Institute of Technology, Chennai
• Sheetal Kumar, Hitachi India R&D
• Krishna Madgula, Zopsmart Technologies
• xOnkar Krishna, Hitachi Ltd, Japan
• Ramesh Kannan R, Vellore Institute of Technology, Chennai
• Saima Mohan, Hitachi India R&D
• Sharath Kumar, Hitachi India R&D
• Yuta Koreeda, Hitachi Ltd, Japan

Paper Submission Instructions

All papers will be double-blind reviewed and must present original work.

• CMT Submission Site
• Select the track: Virtual Special Session 14: Handling Resource constraints for/using ML

Papers submitted for reviewing should conform to IEEE specifications. Manuscript templates can be downloaded from:

• IEEE website

Keydates

• Submission due date: September 5, 2023
• Notification of Acceptance: September 25, 2023
• Camera Ready Papers: October 5, 2023
• Pre-registration: October 15, 2023
• Conference: December 15-17, 2023

Registration

In order for your paper to be presented in the virtual session and published in the proceedings you must register to the conference.

Paper Presentation Instructions

The papers submitted to this track will be presented in the virtual part of the conference. There is no in-person presentation for this session.

If you decide to participate in-person to the conference you must have an adequate internet connection for your presentation and to participate in this special session.