Quantifying the Uncertainty of Human Activity Recognition Using a Bayesian Machine Learning Method: A Prediction Study
Mamiya, H.; Fuller, D.
Show abstract
BackgroundMachine learning methods accurately predict physical activity outcomes using accelerometer data generated by wearable devices, thus allowing the investigation of the impact of built environment on population physical activity. While traditional machine learning methods do not provide prediction uncertainty, a new method, Bayesian Additive Regression Trees (BART) can quantify such uncertainty as posterior predictive distribution. We evaluated the performance of BART in predicting physical activity status. MethodsWe applied multinomial BART and the benchmark method, random forest, to accelerometer data in 25,424 time points, which were generated by wearable devices attached to 37 participants. We evaluated prediction accuracies and confusion matrix using leave-one-out cross-validation. ResultsBART and random forest demonstrated comparable accuracies in prediction. ConclusionsBART is a relatively novel ML method and will advance the incorporation of predicted physical activity status into built environment research. Future research includes the evaluation of the association between the built environment and predicted physical activity with and without accounting for prediction uncertainty.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- From Movement to METs: A Validation of ActTrust(R) for Energy Expenditure Estimation and Physical Activity Classification in Young Adults 94%
- Behaviour-based movement cut-off points in 3-year old children comparing wrist- with hip-worn actigraphs MW8 and GT3X 94%
- Physical determinants of daily physical activity in older men and women 93%
Similar papers in this journal
- Vigorous intermittent lifestyle physical activity (VILPA) and mortality risk among US adults: a wearables-based national cohort study 92%
- Descriptive epidemiology of physical activity energy expenditure in UK adults. The Fenland Study. 90%
- New job, new habits? A Multilevel Interrupted Time Series analysis of diet, physical activity and sleep changes among young adults starting work for the first time 90%
Similar papers in this journal
- Population Analysis Of Mortality Risk: Predictive Models Using Motion Sensors For 100,000 Participants In The UK Biobank National Cohort 91%
- Machine learning-based equations for improved body composition estimation in Indian adults 90%
- How can digital citizen science approaches improve ethical smartphone use surveillance among youth: traditional surveys versus ecological momentary assessments 88%
Similar papers in this journal
- Joint associations between objectively measured physical activity volume and intensity with body-fatness. The Fenland Study 91%
- Evidence for protein leverage on Total Energy Intake, but not Body Mass Index, in a large cohort of older adults 88%
- Association between Accelerometer-Measured Irregular Sleep Duration and Longitudinal Changes in Body Mass Index in Older Adults 87%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.