Predicting stress and burnout levels of remote employees using machine learning and a wearable sensor case study
| dc.contributor.advisor | Maglogiannis, Ilias | |
| dc.contributor.advisor | Μαγκλογιάννης, Ηλίας | |
| dc.contributor.author | Samara, Dimitra | |
| dc.contributor.author | Σαμαρά, Δήμητρα | |
| dc.date.accessioned | 2026-03-31T12:58:56Z | |
| dc.date.available | 2026-03-31T12:58:56Z | |
| dc.date.issued | 2025-12 | |
| dc.identifier.uri | https://dione.lib.unipi.gr/xmlui/handle/unipi/19094 | |
| dc.format.extent | 68 | el |
| dc.language.iso | en | el |
| dc.publisher | Πανεπιστήμιο Πειραιώς | el |
| dc.rights | Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα | * |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/3.0/gr/ | * |
| dc.title | Predicting stress and burnout levels of remote employees using machine learning and a wearable sensor case study | el |
| dc.type | Master Thesis | el |
| dc.contributor.department | Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών Συστημάτων | el |
| dc.description.abstractEN | The current study aims to present a detailed analysis about stress detection using data gathered from chatbot as well as data collected from wearables. The main target is to build a chatbot in order to collect stress related data from employees. Then predictive models such as Random Forest Classifier and XGBoost are used for stress classification as well as for burnout prediction. Both cases are converted to binary classification models using suitable methods. These models are evaluated using robust metrics such as accuracy, precision, F1-score and confusion matrix. Moreover, SMOTE algorithm, which is an algorithm for producing synthetic data, is used in order to achieve class balance and Grid search parameters are used to get the best parameters for the predictive models. In addition to this, the most impactful features that are responsible for the presence of burnout are presented in detail. Regarding the psychological signals collected from wearables, they are used also to build a pipeline for stress classification. In this case, the two predictive models that are used are Random Forest Classifier and Support Vector Machines. These models are also evaluated using the same metrics already mentioned. | el |
| dc.contributor.master | Πληροφοριακά Συστήματα και Υπηρεσίες | el |
| dc.subject.keyword | Machine learning | el |
| dc.subject.keyword | Predicting stress | el |
| dc.subject.keyword | Chatbot | el |
| dc.subject.keyword | Big data | el |
| dc.date.defense | 2025 |
Αρχεία σε αυτό το τεκμήριο
Αυτό το τεκμήριο εμφανίζεται στις ακόλουθες συλλογές
-
Τμήμα Ψηφιακών Συστημάτων
Department of Digital Systems
Εκτός από όπου διευκρινίζεται διαφορετικά, το τεκμήριο διανέμεται με την ακόλουθη άδεια:
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα