GreenML: A Multi-Objective Framework for Energy-Efficient and Sustainable Machine Learning Model Selection
International Journal of Emerging Technologies and Innovative Research · 21.05.2026 · 48
DOI: 10.48175/ijetir-8206 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Emerging Technologies and Innovative Research, online erschienen am 21.05.2026. Verfasst von Dr. Pankaj Agarkar und 3 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben.
Abstract (Originalsprache)
Artificial intelligence and machine learning have become foundational technologies across virtually every industry. Yet the computational demands of developing these systems carry a measurable environmental toll—one that the field has historically chosen to ignore. GreenML is a web- based, multi-objective decision-support framework designed to bridge this gap by integrating environmental accountability directly into the machine learning model selection workflow. The system trains multiple scikit-learn models in parallel on user-supplied tabular datasets, measures their predictive performance alongside training time, energy consumption (kWh), and CO₂-equivalent emissions, and ranks the results using a Pareto- optimal analysis combined with a user-configurable performance-versus-sustainability weighting scheme. Hardware-tier detection further allows the system to issue deployment-aware recommendations tailored to edge or cloud environments. Empirical evaluation on a real-world regression dataset (Uber fare prediction) confirms that lightweight models such as Decision Trees can achieve competitive predictive utility at a fraction of the energy expenditure of ensemble methods, occupying the Pareto front across multiple weighting configurations. GreenML makes the principles of Green AI accessible to everyday practitioners without requiring any background in energy measurement, multi-objective optimization, or sustainability reporting.
Zitieren
APA
Agarkar, D. P., Sonawane, R., Kawde, L., & Suthar, M. (2026). GreenML: A Multi-Objective Framework for Energy-Efficient and Sustainable Machine Learning Model Selection. International Journal of Emerging Technologies and Innovative Research, 48. https://doi.org/10.48175/ijetir-8206
Vancouver
Agarkar DP, Sonawane R, Kawde L, Suthar M. GreenML: A Multi-Objective Framework for Energy-Efficient and Sustainable Machine Learning Model Selection. International Journal of Emerging Technologies and Innovative Research. 2026:48. doi:10.48175/ijetir-8206
BibTeX
@article{Agarkar2026,
author = {Agarkar, Dr. Pankaj and Sonawane, Rudrani and Kawde, Loukik and Suthar, Maheshchand},
title = {GreenML: A Multi-Objective Framework for Energy-Efficient and Sustainable Machine Learning Model Selection},
journal = {International Journal of Emerging Technologies and Innovative Research},
year = {2026},
pages = {48},
doi = {10.48175/ijetir-8206}
}
EUVIMED