The impact of domain shift on predicting perceived sleep quality from wearables
- Abdalazim, Nouran Ayman Roushdy ORCID Facoltà di scienze informatiche, Università della Svizzera italiana, Svizzera
- Alchieri, Leonardo ORCID Facoltà di scienze informatiche, Università della Svizzera italiana, Svizzera
- Alecci, Lidia ORCID Facoltà di scienze informatiche, Università della Svizzera italiana, Svizzera
- Barbiero, Pietro ORCID Facoltà di scienze informatiche, Università della Svizzera italiana, Svizzera
- Santini, Silvia ORCID Facoltà di scienze informatiche, Università della Svizzera italiana, Svizzera
- 2025
Published in:
- Sensors. - 2025, vol. 25, no. 13, p. 4012
English
Machine learning models for personal informatics systems are typically trained offline on records of a specific population of users, resulting in population models. These models may suffer performance degradation in real-world settings due to domain shift, i.e., differences in data distributions across users and contexts. Domain adaptation techniques can address this issue by, e.g., personalizing models with user-specific data. In this paper, we quantify the impact of domain shift on the performance of both population and personalized models in a specific scenario: sleep quality recognition. To this end, we also collect and make available to the research community the new BiheartS dataset. Our analysis shows that domain shift causes the accuracy of population models to decrease by up to 18.54 percentage points, when used on new data. Personalized models, instead, show robust performance across datasets. However, crafting personalized models typically requires using new data or user-provided labels, limiting their applicability in real settings. To mitigate the limitations of both population and personalized models, we propose a novel unsupervised domain adaptation approach: the cluster-based population model (CBPM). CBPM achieves accuracy improvements of up to 13.45 percentage points w.r.t. population model without requiring the use of user-specific records or labels.
- Collections
- Language
-
- English
- Classification
- Computer science and technology
- License
-
CC BY
- Open access status
- gold
- Identifiers
-
- ISSN 1424-8220
- DOI 10.3390/s25134012
- RICERCO 67671
- ARK ark:/12658/srd1335402
- Persistent URL
- https://n2t.net/ark:/12658/srd1335402
Statistics
Document views: 2
File downloads:
- Abdalazim_2025_MDPI_s25134012: 1