Description:
This Reprint brings together recent research on the sustainability of machine learning across critical domains. As AI and smart systems grow, vast data and complex decision-making raise concerns around reliability, efficiency, security, privacy, and societal impact. Contributions examine sustainability from both theoretical and practical perspectives-beyond computational efficiency to include dependability, robustness, ethics, and governance. The works present approaches for trustworthy and resilient ML in real-world settings.
Topics include dependable learning, deep neural network optimization and acceleration, privacy-preserving and federated learning, security in generative models, ethical AI, and sustainability of NLP systems. Applied studies may span healthcare, smart cities, Industry 4.0, sustainable supply chains, and circular economy. Combining methodological and application-driven research, this Reprint offers a valuable reference for researchers, practitioners, and decision-makers focused on sustainable ML in high-impact and mission-critical contexts.