Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

External Research Organisations

  • Delft University of Technology (TU Delft)
  • Wageningen University & Research (WUR)
View graph of relations

Details

Original languageEnglish
Title of host publicationProceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages100-111
Number of pages12
ISBN (electronic)9798331587178
ISBN (print)979-8-3315-8718-5
Publication statusPublished - 9 Jun 2025
Event11th International Conference on ICT for Sustainability, ICT4S 2025 - Dublin, Ireland
Duration: 9 Jun 202513 Jun 2025

Abstract

The reliability of machine learning (ML) software systems is heavily influenced by changes in data over time. For that reason, ML systems require regular maintenance, typically based on model retraining. However, retraining requires significant computational demand, which makes it energy-intensive and raises concerns about its environmental impact. To understand which retraining techniques should be considered when designing sustainable ML applications, in this work, we study the energy consumption of common retraining techniques. Since the accuracy of ML systems is also essential, we compare retraining techniques in terms of both energy efficiency and accuracy. We showcase that retraining with only the most recent data compared to all available data reduces energy consumption by up to 25%, being a sustainable alternative to the status quo. Furthermore, our findings show that retraining a model only when there is evidence that updates are necessary, rather than on a fixed schedule, can reduce energy consumption by up to 40%, provided a reliable data change detector is in place. Our findings pave the way for better recommendations for ML practitioners, guiding them toward more energy-efficient retraining techniques when designing sustainable ML software systems.

Keywords

    Green AI, ML-enabled systems, sustainable AI monitoring, sustainable retraining

ASJC Scopus subject areas

Sustainable Development Goals

Cite this

Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy. / Poenaru-Olaru, Lorena; Sallou, June; Miranda da Cruz, Luis et al.
Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025. Institute of Electrical and Electronics Engineers Inc., 2025. p. 100-111.

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Poenaru-Olaru, L, Sallou, J, Miranda da Cruz, L, Rellermeyer, JS & Van Deursen, A 2025, Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy. in Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025. Institute of Electrical and Electronics Engineers Inc., pp. 100-111, 11th International Conference on ICT for Sustainability, ICT4S 2025, Dublin, Ireland, 9 Jun 2025. https://doi.org/10.1109/ICT4S68164.2025.00019, https://doi.org/10.48550/arXiv.2506.13838
Poenaru-Olaru, L., Sallou, J., Miranda da Cruz, L., Rellermeyer, J. S., & Van Deursen, A. (2025). Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy. In Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025 (pp. 100-111). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICT4S68164.2025.00019, https://doi.org/10.48550/arXiv.2506.13838
Poenaru-Olaru L, Sallou J, Miranda da Cruz L, Rellermeyer JS, Van Deursen A. Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy. In Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025. Institute of Electrical and Electronics Engineers Inc. 2025. p. 100-111 doi: 10.1109/ICT4S68164.2025.00019, 10.48550/arXiv.2506.13838
Poenaru-Olaru, Lorena ; Sallou, June ; Miranda da Cruz, Luis et al. / Sustainable Machine Learning Retraining : Optimizing Energy Efficiency Without Compromising Accuracy. Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025. Institute of Electrical and Electronics Engineers Inc., 2025. pp. 100-111
Download
@inproceedings{4eade2c8a9364df7a2fff0c8be3666aa,
title = "Sustainable Machine Learning Retraining: Optimizing Energy Efficiency Without Compromising Accuracy",
abstract = "The reliability of machine learning (ML) software systems is heavily influenced by changes in data over time. For that reason, ML systems require regular maintenance, typically based on model retraining. However, retraining requires significant computational demand, which makes it energy-intensive and raises concerns about its environmental impact. To understand which retraining techniques should be considered when designing sustainable ML applications, in this work, we study the energy consumption of common retraining techniques. Since the accuracy of ML systems is also essential, we compare retraining techniques in terms of both energy efficiency and accuracy. We showcase that retraining with only the most recent data compared to all available data reduces energy consumption by up to 25%, being a sustainable alternative to the status quo. Furthermore, our findings show that retraining a model only when there is evidence that updates are necessary, rather than on a fixed schedule, can reduce energy consumption by up to 40%, provided a reliable data change detector is in place. Our findings pave the way for better recommendations for ML practitioners, guiding them toward more energy-efficient retraining techniques when designing sustainable ML software systems.",
keywords = "Green AI, ML-enabled systems, sustainable AI monitoring, sustainable retraining",
author = "Lorena Poenaru-Olaru and June Sallou and {Miranda da Cruz}, Luis and Rellermeyer, {Jan S.} and {Van Deursen}, Arie",
note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 11th International Conference on ICT for Sustainability, ICT4S 2025, ICT4S 2025 ; Conference date: 09-06-2025 Through 13-06-2025",
year = "2025",
month = jun,
day = "9",
doi = "10.1109/ICT4S68164.2025.00019",
language = "English",
isbn = "979-8-3315-8718-5",
pages = "100--111",
booktitle = "Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

Download

TY - GEN

T1 - Sustainable Machine Learning Retraining

T2 - 11th International Conference on ICT for Sustainability, ICT4S 2025

AU - Poenaru-Olaru, Lorena

AU - Sallou, June

AU - Miranda da Cruz, Luis

AU - Rellermeyer, Jan S.

AU - Van Deursen, Arie

N1 - Publisher Copyright: © 2025 IEEE.

PY - 2025/6/9

Y1 - 2025/6/9

N2 - The reliability of machine learning (ML) software systems is heavily influenced by changes in data over time. For that reason, ML systems require regular maintenance, typically based on model retraining. However, retraining requires significant computational demand, which makes it energy-intensive and raises concerns about its environmental impact. To understand which retraining techniques should be considered when designing sustainable ML applications, in this work, we study the energy consumption of common retraining techniques. Since the accuracy of ML systems is also essential, we compare retraining techniques in terms of both energy efficiency and accuracy. We showcase that retraining with only the most recent data compared to all available data reduces energy consumption by up to 25%, being a sustainable alternative to the status quo. Furthermore, our findings show that retraining a model only when there is evidence that updates are necessary, rather than on a fixed schedule, can reduce energy consumption by up to 40%, provided a reliable data change detector is in place. Our findings pave the way for better recommendations for ML practitioners, guiding them toward more energy-efficient retraining techniques when designing sustainable ML software systems.

AB - The reliability of machine learning (ML) software systems is heavily influenced by changes in data over time. For that reason, ML systems require regular maintenance, typically based on model retraining. However, retraining requires significant computational demand, which makes it energy-intensive and raises concerns about its environmental impact. To understand which retraining techniques should be considered when designing sustainable ML applications, in this work, we study the energy consumption of common retraining techniques. Since the accuracy of ML systems is also essential, we compare retraining techniques in terms of both energy efficiency and accuracy. We showcase that retraining with only the most recent data compared to all available data reduces energy consumption by up to 25%, being a sustainable alternative to the status quo. Furthermore, our findings show that retraining a model only when there is evidence that updates are necessary, rather than on a fixed schedule, can reduce energy consumption by up to 40%, provided a reliable data change detector is in place. Our findings pave the way for better recommendations for ML practitioners, guiding them toward more energy-efficient retraining techniques when designing sustainable ML software systems.

KW - Green AI

KW - ML-enabled systems

KW - sustainable AI monitoring

KW - sustainable retraining

UR - http://www.scopus.com/inward/record.url?scp=105018907894&partnerID=8YFLogxK

U2 - 10.1109/ICT4S68164.2025.00019

DO - 10.1109/ICT4S68164.2025.00019

M3 - Conference contribution

AN - SCOPUS:105018907894

SN - 979-8-3315-8718-5

SP - 100

EP - 111

BT - Proceedings - 2025 11th International Conference on ICT for Sustainability, ICT4S 2025

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 9 June 2025 through 13 June 2025

ER -

By the same author(s)