Interpretable ML for Stress Detection from Vital Signs Using SHAP

Samson Mihirette; Enrique Antonio De la Cal Martin; Qing Tan

doi:10.1007/978-3-032-08465-1_9

Interpretable ML for Stress Detection from Vital Signs Using SHAP

Samson Mihirette
, Enrique Antonio De la Cal Martin
, Qing Tan

School of Computing and Information systems

University of Oviedo

Research output: Chapter in Book/Report/Conference proceeding › Published Conference contribution › peer-review

1 Citation (Scopus)

Abstract

Stress is a major health concern, contributing to cardiovascular and mental health issues, as well as reduced productivity. Wearable technologies enable continuous monitoring of vital signs—such as heart rate (HR), respiratory rate (RR), and heart rate variability (HRV)—supporting real-time stress detection. Although machine learning (ML) methods have shown strong predictive capabilities, interpretability remains essential for clinical and personal health use. This study integrates SHapley Additive exPlanations (SHAP) with an XGBoost classifier to enhance transparency in stress prediction from physiological signals. Using data from three wearable datasets, SHAP identified HR, RR intervals, and HRV metrics—particularly the LF/HF ratio—as key predictors. The method also revealed the impact of temporal and individual differences on predictions. These findings highlight the potential of SHAP to deliver both accurate and interpretable stress monitoring, advancing trustworthy AI integration in wearable health systems.

Original language	English
Title of host publication	Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings
Editors	Emilio Corchado, Héctor Quintián, Alicia Troncoso Lora, Hilde Pérez García, Esteban Jove Pérez, José Luis Calvo Rolle, Francisco Javier Martínez de Pisón, Pablo García Bringas, Francisco Martínez Álvarez, Álvaro Herrero, Paolo Fosci, Ramos Sérgio Filipe
Pages	103-114
Number of pages	12
DOIs	https://doi.org/10.1007/978-3-032-08465-1_9
Publication status	Published - 2026
Event	20th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2025 - Salamanca, Spain Duration: 16 Oct. 2025 → 17 Oct. 2025

Publication series

Name	Lecture Notes in Computer Science
Volume	16202 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2025
Country/Territory	Spain
City	Salamanca
Period	16/10/25 → 17/10/25

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Machine learning
SHAP
Stress prediction
Vital signs

Access to Document

10.1007/978-3-032-08465-1_9

Cite this

Mihirette, S., De la Cal Martin, E. A., & Tan, Q. (2026). Interpretable ML for Stress Detection from Vital Signs Using SHAP. In E. Corchado, H. Quintián, A. Troncoso Lora, H. Pérez García, E. Jove Pérez, J. L. Calvo Rolle, F. J. Martínez de Pisón, P. García Bringas, F. Martínez Álvarez, Á. Herrero, P. Fosci, & R. Sérgio Filipe (Eds.), Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings (pp. 103-114). (Lecture Notes in Computer Science; Vol. 16202 LNCS). https://doi.org/10.1007/978-3-032-08465-1_9

Mihirette, Samson ; De la Cal Martin, Enrique Antonio ; Tan, Qing. / Interpretable ML for Stress Detection from Vital Signs Using SHAP. Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings. editor / Emilio Corchado ; Héctor Quintián ; Alicia Troncoso Lora ; Hilde Pérez García ; Esteban Jove Pérez ; José Luis Calvo Rolle ; Francisco Javier Martínez de Pisón ; Pablo García Bringas ; Francisco Martínez Álvarez ; Álvaro Herrero ; Paolo Fosci ; Ramos Sérgio Filipe. 2026. pp. 103-114 (Lecture Notes in Computer Science).

@inproceedings{cd4b89e058d0461db7b9d658b7f6b270,

title = "Interpretable ML for Stress Detection from Vital Signs Using SHAP",

abstract = "Stress is a major health concern, contributing to cardiovascular and mental health issues, as well as reduced productivity. Wearable technologies enable continuous monitoring of vital signs—such as heart rate (HR), respiratory rate (RR), and heart rate variability (HRV)—supporting real-time stress detection. Although machine learning (ML) methods have shown strong predictive capabilities, interpretability remains essential for clinical and personal health use. This study integrates SHapley Additive exPlanations (SHAP) with an XGBoost classifier to enhance transparency in stress prediction from physiological signals. Using data from three wearable datasets, SHAP identified HR, RR intervals, and HRV metrics—particularly the LF/HF ratio—as key predictors. The method also revealed the impact of temporal and individual differences on predictions. These findings highlight the potential of SHAP to deliver both accurate and interpretable stress monitoring, advancing trustworthy AI integration in wearable health systems.",

keywords = "Machine learning, SHAP, Stress prediction, Vital signs",

author = "Samson Mihirette and \{De la Cal Martin\}, \{Enrique Antonio\} and Qing Tan",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.; 20th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2025 ; Conference date: 16-10-2025 Through 17-10-2025",

year = "2026",

doi = "10.1007/978-3-032-08465-1\_9",

language = "English",

isbn = "9783032084644",

series = "Lecture Notes in Computer Science",

pages = "103--114",

editor = "Emilio Corchado and H{\'e}ctor Quinti{\'a}n and \{Troncoso Lora\}, Alicia and \{P{\'e}rez Garc{\'i}a\}, Hilde and \{Jove P{\'e}rez\}, Esteban and \{Calvo Rolle\}, \{Jos{\'e} Luis\} and \{Mart{\'i}nez de Pis{\'o}n\}, \{Francisco Javier\} and \{Garc{\'i}a Bringas\}, Pablo and \{Mart{\'i}nez {\'A}lvarez\}, Francisco and {\'A}lvaro Herrero and Paolo Fosci and \{S{\'e}rgio Filipe\}, Ramos",

booktitle = "Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings",

}

Mihirette, S, De la Cal Martin, EA & Tan, Q 2026, Interpretable ML for Stress Detection from Vital Signs Using SHAP. in E Corchado, H Quintián, A Troncoso Lora, H Pérez García, E Jove Pérez, JL Calvo Rolle, FJ Martínez de Pisón, P García Bringas, F Martínez Álvarez, Á Herrero, P Fosci & R Sérgio Filipe (eds), Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings. Lecture Notes in Computer Science, vol. 16202 LNCS, pp. 103-114, 20th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2025, Salamanca, Spain, 16/10/25. https://doi.org/10.1007/978-3-032-08465-1_9

Interpretable ML for Stress Detection from Vital Signs Using SHAP. / Mihirette, Samson; De la Cal Martin, Enrique Antonio; Tan, Qing.
Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings. ed. / Emilio Corchado; Héctor Quintián; Alicia Troncoso Lora; Hilde Pérez García; Esteban Jove Pérez; José Luis Calvo Rolle; Francisco Javier Martínez de Pisón; Pablo García Bringas; Francisco Martínez Álvarez; Álvaro Herrero; Paolo Fosci; Ramos Sérgio Filipe. 2026. p. 103-114 (Lecture Notes in Computer Science; Vol. 16202 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Published Conference contribution › peer-review

TY - GEN

T1 - Interpretable ML for Stress Detection from Vital Signs Using SHAP

AU - Mihirette, Samson

AU - De la Cal Martin, Enrique Antonio

AU - Tan, Qing

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.

PY - 2026

Y1 - 2026

N2 - Stress is a major health concern, contributing to cardiovascular and mental health issues, as well as reduced productivity. Wearable technologies enable continuous monitoring of vital signs—such as heart rate (HR), respiratory rate (RR), and heart rate variability (HRV)—supporting real-time stress detection. Although machine learning (ML) methods have shown strong predictive capabilities, interpretability remains essential for clinical and personal health use. This study integrates SHapley Additive exPlanations (SHAP) with an XGBoost classifier to enhance transparency in stress prediction from physiological signals. Using data from three wearable datasets, SHAP identified HR, RR intervals, and HRV metrics—particularly the LF/HF ratio—as key predictors. The method also revealed the impact of temporal and individual differences on predictions. These findings highlight the potential of SHAP to deliver both accurate and interpretable stress monitoring, advancing trustworthy AI integration in wearable health systems.

AB - Stress is a major health concern, contributing to cardiovascular and mental health issues, as well as reduced productivity. Wearable technologies enable continuous monitoring of vital signs—such as heart rate (HR), respiratory rate (RR), and heart rate variability (HRV)—supporting real-time stress detection. Although machine learning (ML) methods have shown strong predictive capabilities, interpretability remains essential for clinical and personal health use. This study integrates SHapley Additive exPlanations (SHAP) with an XGBoost classifier to enhance transparency in stress prediction from physiological signals. Using data from three wearable datasets, SHAP identified HR, RR intervals, and HRV metrics—particularly the LF/HF ratio—as key predictors. The method also revealed the impact of temporal and individual differences on predictions. These findings highlight the potential of SHAP to deliver both accurate and interpretable stress monitoring, advancing trustworthy AI integration in wearable health systems.

KW - Machine learning

KW - SHAP

KW - Stress prediction

KW - Vital signs

UR - https://www.scopus.com/pages/publications/105020678006

U2 - 10.1007/978-3-032-08465-1_9

DO - 10.1007/978-3-032-08465-1_9

M3 - Published Conference contribution

AN - SCOPUS:105020678006

SN - 9783032084644

T3 - Lecture Notes in Computer Science

SP - 103

EP - 114

BT - Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings

A2 - Corchado, Emilio

A2 - Quintián, Héctor

A2 - Troncoso Lora, Alicia

A2 - Pérez García, Hilde

A2 - Jove Pérez, Esteban

A2 - Calvo Rolle, José Luis

A2 - Martínez de Pisón, Francisco Javier

A2 - García Bringas, Pablo

A2 - Martínez Álvarez, Francisco

A2 - Herrero, Álvaro

A2 - Fosci, Paolo

A2 - Sérgio Filipe, Ramos

T2 - 20th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2025

Y2 - 16 October 2025 through 17 October 2025

ER -

Mihirette S, De la Cal Martin EA, Tan Q. Interpretable ML for Stress Detection from Vital Signs Using SHAP. In Corchado E, Quintián H, Troncoso Lora A, Pérez García H, Jove Pérez E, Calvo Rolle JL, Martínez de Pisón FJ, García Bringas P, Martínez Álvarez F, Herrero Á, Fosci P, Sérgio Filipe R, editors, Hybrid Artificial Intelligent Systems - 20th International Conference, HAIS 2025, Proceedings. 2026. p. 103-114. (Lecture Notes in Computer Science). doi: 10.1007/978-3-032-08465-1_9