Data analytic approaches to detect stress from physiological data
- Publication Type:
- Thesis
- Issue Date:
- 2025
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
Stress has become one of the most critical dangers to human health. Recently, with the development of wearable devices that can measure physiological features more accurately and collect more data, stress detection using the data collected from wearable devices has become an emerging research topic. There are currently many studies on detecting and predicting stress. However, these studies are not comprehensive enough because (i) they provide limited comparison with other methods, (ii) they simply use statistics to show correlation between physiological features, (iii) they use mainly classical methods or advanced methods with raw physiological data, or (iv) they do not compare to other methods using benchmark datasets. New data analytic approaches, such as topological data analysis (TDA), which is robust to noise and can analyse high-dimensional data, can also create more opportunities for research about stress. Using complex networks representations of the physiological data and associated algorithms to detect stress is a promising solution since complex networks have the potential to capture phenomena of real-world systems. This thesis presents novel approaches to present the relationship between stress and physiological time series features using topological data analysis and complex networks. Benchmark datasets are used to deploy experiments of this thesis.
Please use this identifier to cite or link to this item: