Call for Papers
Special Issue on Sustainable AI: Methods and Scientific Applications
This Special issue marks the inception of Sustainable Artificial Intelligence (Sustainable AI) literature, exploring its profound research potential by harnessing computational resources more efficiently and extending its applications across a spectrum of scientific disciplines. In our contemporary world, the imperative to establish a sustainable environment for future generations underscores the critical need for Sustainable AI. It serves as a strategic lens through which we can evaluate the global impact of AI on climate change. The ripple effect of Sustainable AI extends to the facilitation of emerging scientific applications, such as solar cells, catalyst screening, semiconductor failure analysis and low-carbon technologies.
The primary objective of this Special issue is to consolidate the latest breakthroughs in Sustainable AI research, encompassing both algorithmic advancements and scientific applications. The focus is centered on four key themes: 1) Data-Efficient Algorithms, 2) Resource-Efficient Algorithms, 3) Hybrid (Data & Resource Efficient) Algorithms, and 4) Emerging Scientific applications.
By delving into these topics, this special issue not only contributes to the academic discourse surrounding Sustainable AI but also highlights its practical implications in fostering a sustainable future and advancing scientific frontiers.
Topics of interest include, but are not limited to, the following:
- I: Introduction to Sustainable AI
Sparse data regime
Learning-based methods
Optimization techniques
Efficient learning - II: Data-efficient learning
Active learning
Semi-supervised learning
Foundational models
Prompt-based learning
Transfer Learning
Domain adaptation
Domain generalization - III: Resource-efficient learning
Self-evolving architecture
Online learning
Online continual learning
Knowledge Distillation
Network pruning and quantization
Bayesian optimization - IV: Hybrid-efficient learning
Evolving unsupervised representation
Autonomous deep model compression
Physics-embedded model compression
Kalman filter embedded recurrent learning - Sustainable AI for Scientific Applications
Sustainable AI for catalyst screening
Sustainable AI for solar cells
Sustainable AI for semiconductor failure analysis
Sustainable AI for low carbon technologies
Important Deadlines:
- Submission deadline: 31 July 2024
Publication date: 30 November 2024
Submission Instructions:
Please read the [Guide for Authors] before submitting.
All articles should be [submitted online], please select Article Type: Sustainable AI: Methods and Scientific Applications on submission.
Guest Editors:
J Senthilnath Institute for Infocomm Research, A*STAR, Singapore
Xiaoli Li Institute for Infocomm Research, A*STAR; Nanyang Technological University, Singapore
Yung-Hsiang Lu Electrical and Computer Engineering, Purdue University, Indiana, USA