Unsupervised Classification of Human Activity with Hidden Semi-Markov Models
The modern sedentary lifestyle is negatively influencing human health, and the current guidelines recommend at least 150 min of moderate activity per week. However, the challenge is how to measure human activity in a practical way. While accelerometers are the most common tools to measure activity,...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Applied System Innovation |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2571-5577/5/4/83 |
| _version_ | 1797439579141177344 |
|---|---|
| author | Francesca Romana Cavallo Christofer Toumazou Konstantin Nikolic |
| author_facet | Francesca Romana Cavallo Christofer Toumazou Konstantin Nikolic |
| author_sort | Francesca Romana Cavallo |
| collection | DOAJ |
| description | The modern sedentary lifestyle is negatively influencing human health, and the current guidelines recommend at least 150 min of moderate activity per week. However, the challenge is how to measure human activity in a practical way. While accelerometers are the most common tools to measure activity, current activity classification methods require calibration studies or labelled datasets—requirements that slow the research progress. Therefore, there is a pressing need to classify and quantify human activity efficiently. In this work, we propose an unsupervised approach to classify activities from accelerometer data using hidden semi-Markov models. We tune and infer the model parameters on accelerometer data from the UK Biobank and select the optimal model based on features used and informativeness of the prior. The best model achieves an average correlation of 0.4 between the inferred activities and the reference ones, with the overall physical activity obtaining a correlation of 0.8. Additionally, to prove the clinical significance of the method, we validate it by performing a linear regression between the inferred activities and anthropometric measures such as BMI and waist circumference. We show that for a sedentary behaviour and total physical activity, the proposed method achieves comparable regression coefficients to the reference labelled dataset. Moreover, the proposed method achieves a good agreement with a labelled dataset for daily time spent in a sedentary behaviour and total physical activity. The unsupervised nature of the method allows for a data-driven classification that does not require calibration studies or labelled datasets and can thus facilitate both clinical research as well as lifestyle recommendations. |
| first_indexed | 2024-03-09T11:56:07Z |
| format | Article |
| id | doaj.art-f5eab5fe3f394aa68085ff2f659bcf7a |
| institution | Directory Open Access Journal |
| issn | 2571-5577 |
| language | English |
| last_indexed | 2024-03-09T11:56:07Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied System Innovation |
| spelling | doaj.art-f5eab5fe3f394aa68085ff2f659bcf7a2023-11-30T23:09:50ZengMDPI AGApplied System Innovation2571-55772022-08-01548310.3390/asi5040083Unsupervised Classification of Human Activity with Hidden Semi-Markov ModelsFrancesca Romana Cavallo0Christofer Toumazou1Konstantin Nikolic2Centre for Bio-Inspired Technology, Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2AZ, UKCentre for Bio-Inspired Technology, Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2AZ, UKCentre for Bio-Inspired Technology, Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2AZ, UKThe modern sedentary lifestyle is negatively influencing human health, and the current guidelines recommend at least 150 min of moderate activity per week. However, the challenge is how to measure human activity in a practical way. While accelerometers are the most common tools to measure activity, current activity classification methods require calibration studies or labelled datasets—requirements that slow the research progress. Therefore, there is a pressing need to classify and quantify human activity efficiently. In this work, we propose an unsupervised approach to classify activities from accelerometer data using hidden semi-Markov models. We tune and infer the model parameters on accelerometer data from the UK Biobank and select the optimal model based on features used and informativeness of the prior. The best model achieves an average correlation of 0.4 between the inferred activities and the reference ones, with the overall physical activity obtaining a correlation of 0.8. Additionally, to prove the clinical significance of the method, we validate it by performing a linear regression between the inferred activities and anthropometric measures such as BMI and waist circumference. We show that for a sedentary behaviour and total physical activity, the proposed method achieves comparable regression coefficients to the reference labelled dataset. Moreover, the proposed method achieves a good agreement with a labelled dataset for daily time spent in a sedentary behaviour and total physical activity. The unsupervised nature of the method allows for a data-driven classification that does not require calibration studies or labelled datasets and can thus facilitate both clinical research as well as lifestyle recommendations.https://www.mdpi.com/2571-5577/5/4/83activity classificationaccelerometerhidden Markov modelswearable sensorsUK Biobank |
| spellingShingle | Francesca Romana Cavallo Christofer Toumazou Konstantin Nikolic Unsupervised Classification of Human Activity with Hidden Semi-Markov Models Applied System Innovation activity classification accelerometer hidden Markov models wearable sensors UK Biobank |
| title | Unsupervised Classification of Human Activity with Hidden Semi-Markov Models |
| title_full | Unsupervised Classification of Human Activity with Hidden Semi-Markov Models |
| title_fullStr | Unsupervised Classification of Human Activity with Hidden Semi-Markov Models |
| title_full_unstemmed | Unsupervised Classification of Human Activity with Hidden Semi-Markov Models |
| title_short | Unsupervised Classification of Human Activity with Hidden Semi-Markov Models |
| title_sort | unsupervised classification of human activity with hidden semi markov models |
| topic | activity classification accelerometer hidden Markov models wearable sensors UK Biobank |
| url | https://www.mdpi.com/2571-5577/5/4/83 |
| work_keys_str_mv | AT francescaromanacavallo unsupervisedclassificationofhumanactivitywithhiddensemimarkovmodels AT christofertoumazou unsupervisedclassificationofhumanactivitywithhiddensemimarkovmodels AT konstantinnikolic unsupervisedclassificationofhumanactivitywithhiddensemimarkovmodels |