Joana Chong, Petra Tjurin, Maisa Niemelä, Timo Jämsä, Vahid Farrahi, Machine-learning models for activity class prediction: A comparative study of feature selection and classification algorithms, Gait & Posture, Volume 89, 2021, Pages 45-53, ISSN 0966-6362, https://doi.org/10.1016/j.gaitpost.2021.06.017
Machine-learning models for activity class prediction : a comparative study of feature selection and classification algorithms
|Author:||Chong, Joana1,2; Tjurin, Petra2; Niemelä, Maisa2,3;|
1Faculty of Sciences, University of Lisbon, Lisbon, Portugal
2Research Unit of Medical Imaging, Physics and Technology, University of Oulu, Oulu, Finland
3Medical Research Center, Oulu University Hospital and University of Oulu, Oulu, Finland
4Diagnostic Radiology, Oulu University Hospital, Oulu, Finland
|Online Access:||PDF Full Text (PDF, 3.2 MB)|
|Persistent link:|| http://urn.fi/urn:nbn:fi-fe2021111154662
|Publish Date:|| 2021-11-11
Purpose: Machine-learning (ML) approaches have been repeatedly coupled with raw accelerometry to classify physical activity classes, but the features required to optimize their predictive performance are still unknown. Our aim was to identify appropriate combination of feature subsets and prediction algorithms for activity class prediction from hip-based raw acceleration data.
Methods: The hip-based raw acceleration data collected from 27 participants was split into training (70 %) and validation (30 %) subsets. A total of 206 time- (TD) and frequencydomain (FD) features were extracted from 6-second non-overlapping windows of the signal. Feature selection was done using seven filter-based, two wrapper-based, and one embedded algorithm, and classification was performed with artificial neural network (ANN), support vector machine (SVM), and random forest (RF). For every combination between the feature selection method and the classifiers, the most appropriate feature subsets were found and used for model training within the training set. These models were then validated with the left-out validation set.
Results: The appropriate number of features for the ANN, SVM, and RF ranged from 20 to 45. Overall, the accuracy of all the three classifiers was higher when trained with feature subsets generated using filter-based methods compared with when they were trained with wrapper-based methods (range: 78.1 %–88 % vs. 66 %–83.5 %). TD features that reflect how signals vary around the mean, how they differ with one another, and how much and how often they change were more frequently selected via the feature selection methods.
Conclusions: A subset of TD features from raw accelerometry could be sufficient for ML-based activity classification if properly selected from different axes.
Gait & posture
|Pages:||45 - 53|
|Type of Publication:||
A1 Journal article – refereed
|Field of Science:||
315 Sport and fitness sciences
217 Medical engineering
This work was supported in part by the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement no. 713645, the Ministry of Education and Culture in Finland (OKM/86/626/2014, OKM/43/626/2015, OKM/17/626/2016, OKM/54/626/2019, OKM/85/626/2019, OKM/1096/626/2020), and by Infotech Oulu, Finland. The original data collection was supported in part by Polar Electro Oy and the Finnish Funding Agency for Innovation (6057/31/2016).
© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).