Benchmarking energy efficiency of supervised machine learning models on multi-domain classification dataset
File(s)
Published version
Author(s)
Ali, Amir
Qamar, Rohail
Asif, Raheela
Hina, Saman
Type
Journal Article
Abstract
Machine learning should be judged by how well it predicts, and compute is not accounted for in predictive accuracy. Given the growing emphasis on energy consumption and resource efficiency, decision-supporting frameworks should go beyond accuracy.
This study presents an energy-based benchmarking approach for supervised learning models. Ten classical algorithms were evaluated on three textual and tabular datasets. The energy consumption of preprocessing, training, and inference was monitored with Intel RAPL via pyRAPL along with the runtime, peak memory usage, and predictive performance statistics (accuracy, precision, recall, F1-score, and AUC). Experiments were
conducted in a controlled CPU-based environment to ensure comparability. The computational role of this feature is found to be appreciably diverse. Results show that Random Forest achieved the highest overall balance between predictive performance and effi
ciency (CI = 0.950, PPI = 0.907), while Logistic Regression provided a competitive tradeoff (CI = 0.905, EI = 0.998). Gaussian Naïve Bayes was the most energy-efficient model with a mean energy consumption of 127 J, whereas Support Vector Classifier (SVC) incurred
the highest computational cost, consuming 45,758 J and requiring 3,925 s on average. The Pareto analysis identified Random Forest, Logistic Regression, Passive Aggressive, and Decision Tree as non-dominated solutions. These findings demonstrate that accuracy
alone can be misleading for model evaluation and that integrating energy, runtime, and memory metrics enables more sustainable and resource-aware machine learning model selection. The proposed framework provides practical guidance for Green AI, Tiny Ma
chine Learning (TinyML), edge computing, and other resource-constrained deployment
environments.
This study presents an energy-based benchmarking approach for supervised learning models. Ten classical algorithms were evaluated on three textual and tabular datasets. The energy consumption of preprocessing, training, and inference was monitored with Intel RAPL via pyRAPL along with the runtime, peak memory usage, and predictive performance statistics (accuracy, precision, recall, F1-score, and AUC). Experiments were
conducted in a controlled CPU-based environment to ensure comparability. The computational role of this feature is found to be appreciably diverse. Results show that Random Forest achieved the highest overall balance between predictive performance and effi
ciency (CI = 0.950, PPI = 0.907), while Logistic Regression provided a competitive tradeoff (CI = 0.905, EI = 0.998). Gaussian Naïve Bayes was the most energy-efficient model with a mean energy consumption of 127 J, whereas Support Vector Classifier (SVC) incurred
the highest computational cost, consuming 45,758 J and requiring 3,925 s on average. The Pareto analysis identified Random Forest, Logistic Regression, Passive Aggressive, and Decision Tree as non-dominated solutions. These findings demonstrate that accuracy
alone can be misleading for model evaluation and that integrating energy, runtime, and memory metrics enables more sustainable and resource-aware machine learning model selection. The proposed framework provides practical guidance for Green AI, Tiny Ma
chine Learning (TinyML), edge computing, and other resource-constrained deployment
environments.
Date Issued
2026-07-04
Date Acceptance
2026-07-02
Citation
Information, 2026, 17 (7)
ISSN
2078-2489
Publisher
MDPI
Journal / Book Title
Information
Volume
17
Issue
7
Copyright Statement
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.
License URL
Identifier
10.3390/info17070652
Subjects
sustainable AI
machine learning
energy efficient
computational efficiency
model benchmarking
predictive analytics
explainable AI
Green AI
TinyML
pareto optimization
Publication Status
Published
Article Number
ARTN 652