ChatPPG.
Join free

Paperdoi:10.3390/s26216996

Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks

Venue listed2026Added Nov 5, 2026

  • DO
  • LV
  • HF

Authors: Daniela Oliveira, Lucas Veras, Hélder Fonseca

Sensors 26(21), 6996 (published 2026-11-05)

Publisher (opens in a new tab) Cite
CC BY 4.0

TL;DR

Forty-nine adults (28.4 ± 8.5 y) did seven laboratory activities wearing ActiGraph GT9X Link accelerometers on the right hip, dominant wrist, and non-dominant wrist. Energy expenditure from indirect calorimetry. Time- and frequency-domain features from non-overlapping 1-s windows plus age, sex, BMI and body mass.

  • Form factor: wrist

Abstract

Accelerometers are widely used to estimate energy expenditure (EE) and classify physical activity intensity (PAI), although conventional approaches may not capture the complexity of raw acceleration signals. This study evaluated machine learning (ML) models for estimating EE and classifying PAI during daily and locomotor tasks. Forty-nine adults (28.4 ± 8.5 years) completed seven laboratory activities while wearing ActiGraph GT9X Link accelerometers on the right hip and dominant and non-dominant wrists. EE was measured by indirect calorimetry. Time- and frequency-domain accelerometry features from non-overlapping 1-s windows were combined with age, sex, body mass index (BMI), and body mass. K-nearest neighbors (KNN) and random forest (RF) models were developed for each placement using an 80/20 training–test split. RF achieved lower overall EE estimation errors and higher overall PAI classification agreement than KNN across placements. The lowest EE errors were obtained from RF models at the non-dominant wrist (RMSE = 0.67 kcal·min−1; MAE = 0.39 kcal·min−1) and dominant wrist (RMSE = 0.68 kcal·min−1; MAE = 0.40 kcal·min−1). RF classification agreement reached 95.81% (Kappa = 0.943) at the non-dominant wrist and 94.97% (Kappa = 0.931) at the hip. In this laboratory protocol, the non-dominant wrist RF model showed a favorable combination of EE estimation error and PAI classification agreement.

Verbatim from the publisher's DOI record (opens in a new tab) · Publisher licence: https://creativecommons.org/licenses/by/4.0/

Results on ChatPPG leaderboards · 2

All leaderboards
LeaderboardMethodResultRank
oliveira-2026-lab-actigraph mae_kcal_per_minEnergy expenditure (PAEE / calorimetry)Accelerometry-Based Machine Learning Estimation…0.39mae_kcal_per_min, lower is better#1 of 1
oliveira-2026-lab-actigraph accuracyHuman activity recognitionAccelerometry-Based Machine Learning Estimation…0.9581accuracy, higher is better#1 of 1

2 rows are credited to this paper. Ranks apply within each leaderboard only, as reported by the source.

Tasks & signals

Tasks

Energy expenditure (PAEE / calorimetry)Human activity recognitionActivity recognition

Signals

IMU

Keywords

Wrist AccelerometerEnergy ExpenditurePhysical Activity IntensityHARActiGraph GT9XRandom ForestLab

How it works

How Method works: Setup → Method → TargetSetup: Signals (IMU); Form factor (wrist). Method: Method. Target: Activity recognition.SetupSignals: IMUxyzSignalsIMUForm factor: wristForm factorwristMethodMethodTargetActivity recognitionActivity recognition

Cite this paper

Cite this paper with the fields recorded here. Check it against the publisher before submitting.

Oliveira, D., Veras, L., & Fonseca, H. (2026). Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks. Sensors 26, 6996. https://doi.org/10.3390/s26216996
Oliveira, Daniela, Lucas Veras, and Hélder Fonseca. 2026. "Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks." Sensors 26, 6996. https://doi.org/10.3390/s26216996.
Oliveira, D., Veras, L. and Fonseca, H. (2026) 'Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks', Sensors 26, 6996. doi: 10.3390/s26216996.
Oliveira, Daniela, et al. "Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks." Sensors 26, 6996, 2026. https://doi.org/10.3390/s26216996.
D. Oliveira, L. Veras, and H. Fonseca, "Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks," Sensors 26, 6996, 2026, doi: 10.3390/s26216996.
@article{oliveira2026accelerometrybased,
  title         = {{Accelerometry-Based Machine Learning Estimation of Energy Expenditure and Physical Activity Intensity Classification in Daily Tasks}},
  author        = {Oliveira, Daniela and Veras, Lucas and Fonseca, Hélder},
  year          = {2026},
  journal       = {Sensors 26, 6996},
  doi           = {10.3390/s26216996}
}

Added to the ChatPPG research hub

Same tasks, data and authors

Papers

Datasets