Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Oct 22, 2025

Assessment of Physical Activity Intensity with Accelerometers and Oxygen Consumption
08:45

Assessment of Physical Activity Intensity with Accelerometers and Oxygen Consumption

Published on: June 20, 2025

257

Feature selection for unsupervised machine learning of accelerometer data physical activity clusters - A systematic

Petra J Jones1, Mike Catt2, Melanie J Davies3

  • 1Leicester Diabetes Centre, University Hospitals of Leicester, Leicester General Hospital, Gwendolen Road, Leicester LE5 4PW, UK; Diabetes Research Centre, University of Leicester, Leicester General Hospital, Gwendolen Road, Leicester LE5 4PW, UK.

Gait & Posture
|August 26, 2021
PubMed

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Less Is More? Physical Inactivity and Increased Risk of Diabetes-Related Foot Ulceration-A Systematic Review.

Diabetes/metabolism research and reviews·2026
Same author

Is it time to integrate the 24-hour movement paradigm into metabolic and bariatric surgery practice and research? A call to action.

Surgery for obesity and related diseases : official journal of the American Society for Bariatric Surgery·2026
Same author

The Effect of Acute Aerobic Exercise On the Time Spent in Hypoglycaemia After Bariatric Surgery (The BariEX Study).

Obesity surgery·2026
Same author

A Sedentary Behaviour Reduction Programme in Patients With Peripheral Arterial Disease: A Mixed-Method Feasibility Study.

International journal of vascular medicine·2026
Same author

The Utility of Measures of Physical Behavior, Function, and Fitness as Predictors of Mortality.

Mayo Clinic proceedings. Innovations, quality & outcomes·2026
Same author

Interactive deep learning for myocardial scar segmentation using cardiovascular magnetic resonance.

Journal of cardiovascular magnetic resonance : official journal of the Society for Cardiovascular Magnetic Resonance·2026
Summary
This summary is machine-generated.

This review summarizes feature selection for unsupervised machine learning of physical activity (PA) from accelerometers. Principal Component Analysis (PCA) and correlation-based methods were most popular, highlighting a need for larger datasets and clearer reporting.

Area of Science:

  • Biomedical Engineering
  • Data Science
  • Public Health

Background:

  • Physical activity (PA) monitoring using accelerometers is crucial for understanding health risks and promoting healthy behaviors.
  • Unsupervised machine learning offers a way to analyze PA from free-living accelerometer data without requiring labeled datasets.
  • Limited research exists on feature selection techniques for accelerometer-based PA analysis.

Purpose of the Study:

  • To systematically review feature selection techniques used in unsupervised machine learning for accelerometer-based physical activity.
  • To identify commonly used features selected through these techniques.
  • To inform future research on optimizing PA clustering from accelerometer data.

Main Methods:

  • A systematic literature search was conducted across major scientific databases (PubMed, Scopus, Web of Science, etc.) for studies published before January 2021.
Keywords:
AccelerometerClusteringFeature selectionK-meansPhysical activity

More Related Videos

Home-Based Monitor for Gait and Activity Analysis
07:24

Home-Based Monitor for Gait and Activity Analysis

Published on: August 8, 2019

6.9K
Physical Activity Measurement in Children Accepting Table Tennis Training
06:51

Physical Activity Measurement in Children Accepting Table Tennis Training

Published on: July 27, 2022

2.1K

Related Experiment Videos

Last Updated: Oct 22, 2025

Assessment of Physical Activity Intensity with Accelerometers and Oxygen Consumption
08:45

Assessment of Physical Activity Intensity with Accelerometers and Oxygen Consumption

Published on: June 20, 2025

257
Home-Based Monitor for Gait and Activity Analysis
07:24

Home-Based Monitor for Gait and Activity Analysis

Published on: August 8, 2019

6.9K
Physical Activity Measurement in Children Accepting Table Tennis Training
06:51

Physical Activity Measurement in Children Accepting Table Tennis Training

Published on: July 27, 2022

2.1K
  • Studies employing feature selection methods for unsupervised physical activity clustering from accelerometer data were included.
  • Included studies were analyzed for feature selection techniques, commonly selected features, and clustering algorithms.

    • Main Results:

    • Thirteen studies met the inclusion criteria for the review.
    • Principal Component Analysis (PCA) and correlation-based methods were the most frequently used feature selection techniques.
    • K-means was the predominant clustering algorithm, and cluster evaluation methods varied widely.

    Conclusions:

    • There is a need for studies to evaluate multiple feature selection methods on large datasets with multiple physical activity (PA) datasets.
    • Clear reporting of feature selection cut-off criteria (e.g., for PCA) and clustering hyperparameters is essential.
    • Future research should focus on robust feature selection and evaluation for improved PA clustering.