Evaluation of Stress Detection as Time Series Events -- A Novel Window-Based F1-Metric
Preprint
Abstract
Accurate evaluation of event detection in time series is essential for applications such as stress monitoring with wearable devices, where ground truth is typically annotated as single-point events, even though the underlying phenomena are gradual and temporally diffused. Standard metrics like F1 and point-adjusted F1 (F1$_{pa}$) often misrepresent model performance in such real-world, imbalanced datasets. We introduce a window-based F1 metric (F1$_w$) that incorporates temporal tolerance, enabling a more robust assessment of event detection when exact alignment is unrealistic. Empirical analysis in three physiological datasets, two in-the-wild (ADARP, Wrist Angel) and one experimental (ROAD), indicates that F1$_w$ reveals meaningful model performance patterns invisible to conventional metrics, while its window size can be adapted to domain knowledge to avoid overestimation. We show that the choice of evaluation metric strongly influences the interpretation of model performance: using predictions from TimesFM, only our temporally tolerant metrics reveal statistically significant improvements over random and null baselines in the two in-the-wild use cases. This work addresses key gaps in time series evaluation and provides practical guidance for healthcare applications where requirements for temporal precision vary by context.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- Online Event Detection in Streaming Time Series: Novel Metrics and Practical Insights2024
- ADARP: A Multi Modal Dataset for Stress and Alcohol Relapse Quantification in Real Life Setting2022
- A multimodal sensor dataset for continuous stress detection of nurses in a hospitalScientific Data, 2022
- Predicting Driver Self-Reported Stress by Analyzing the Road Scene2021
- An Evaluation of Anomaly Detection and Diagnosis in Multivariate Time SeriesIEEE Transactions on Neural Networks and Learning Systems, 2021
- AffectiveROAD system and database to assess driver's attention2018
- Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications2018
- Performing event detection in time series with SwiftEvent: an algorithm with supervised learning of detection criteriaPattern Analysis and Applications, 2017
- A note on using the F-measure for evaluating record linkage algorithmsStatistics and Computing, 2017
- Event detection from time series data1999