Tools
Tools
Guerrache, F 
ORCID: https://orcid.org/0000-0001-7424-6562,
2024.
Spatial and temporal environment impact analysis on people’s well-being.
PhD, Nottingham Trent University.
Preview |
Text
Faiza Guerrache 2024.pdf - Published version Download (29MB) | Preview |
Abstract
This PhD research programme presents an innovative approach to understand the environmental factors of human wellbeing through the development and analysis of the ”EnviroWellBeing” dataset and the application of advanced machine learning and deep learning algorithms for environment and stress-level classification tasks. The careful curation of the dataset involved the synchronisation of sensor data to a uniform 1Hz frequency and the application of comprehensive data cleaning processes, ensuring its suitability for time-series analysis. The dataset curation effort marks a significant advancement in studying the spatial and temporal impacts of environmental factors on physiological and psychological states.
Additionally, this research explores the ”Depresjon” dataset, applying data analysis techniques to uncover patterns in motor activity related to depression. A comparative analysis of machine learning models demonstrated the ability to distinguish between depressed and healthy individuals using motor activity, with a Random Forest (RF) classifier achieving 83.41% accuracy. Analysis of the Depresjon dataset reveals key physiological markers of depression and highlights the role of predictive modeling in advancing mental health research.
Using the EnviroWellBeing Dataset, the research details the higher performance of 1D Convolutional Neural Networks (1D-CNNs), which achieved notable accuracies in classifying environmental conditions (e.g., 97.72% in situ, 94.18% in vitro) and stress levels (e.g., 82.37% in situ, 63.37% in vitro), highlighting their effectiveness. The evaluation also includes the effectiveness of Long Short-Term Memory (LSTM) networks in capturing sequential dependencies and the robustness of RF classifiers as a non-sequential baseline. Key findings demonstrate the critical role of feature selection (identifying CO2, wrist temperature, and NO as key predictors in situ), the models’ capacity to generalise from physiological responses to stress, and the provision of valuable insights into feature importance for future model development.
By offering valuable insights into the performance of machine learning models in environmental and stress-level classifications, alongside a comprehensive dataset, this PhD research programme significantly contributes to the fields of environmental health and mental wellbeing. The findings demonstrate potential applications in urban planning and personal health monitoring by showing how sensor data analysis can inform strategies to mitigate environmental stressors and enhance human health and happiness.
| Item Type: | Thesis |
|---|---|
| Creators: | Guerrache, F. |
| Contributors: | Name Role NTU ID ORCID |
| Date: | February 2024 |
| Rights: | The copyright in this work is held by the author. You may copy up to 5% of this work for private study, or personal, non-commercial research. Any re-use of the information contained within this document should be fully referenced, quoting the author, title, university, degree level and pagination. Queries or requests for any other use, or if a more substantial copy is required, should be directed to the author. |
| Divisions: | Schools > School of Science and Technology |
| Record created by: | Jeremy Silvester |
| Date Added: | 12 Sep 2025 11:03 |
| Last Modified: | 12 Sep 2025 11:03 |
| URI: | https://irep.ntu.ac.uk/id/eprint/54322 |
Actions (login required)
 |
Edit View |
Statistics
Views
Views per month over past year
Downloads
Downloads per month over past year
