TY - GEN
T1 - Daily stress recognition system using activity tracker and smartphone based on physical activity and heart rate data
AU - Lawanont, Worawat
AU - Mongkolnam, Pornchai
AU - Nukoolkit, Chakarida
AU - Inoue, Masahiro
N1 - Funding Information:
supported by JSPS KAKENHI Grant number
Funding Information:
This work was supported by JSPS KAKENHI Grant number 15K00929.
PY - 2019
Y1 - 2019
N2 - Everyday, people experience stress, and it has been suggested for a long time that stress will eventually develop into anxiety as well as other physical issues. The emerging technology, such as wearable sensors and smartphone, have enabled the opportunity of using the technology to help solve the issue. In this paper, we proposed a system using Internet of Things architecture where we adopted an activity tracker as our sensing device to reduce cumbersome for daily use. Among the total of 17 features extracted from activity tracker, five features from sleep data and six features from heart rate data were proposed to develop the stress recognition model. In the evaluation of our system, we achieved the accuracy as high as 81.70% on the cross validation and 78.95% when tested on the test set. Despite that this is a preliminary result, it has shown that it is possible to use the IoT architecture along with the activity tracker to accurately recognize stress and help improve one’s wellbeing.
AB - Everyday, people experience stress, and it has been suggested for a long time that stress will eventually develop into anxiety as well as other physical issues. The emerging technology, such as wearable sensors and smartphone, have enabled the opportunity of using the technology to help solve the issue. In this paper, we proposed a system using Internet of Things architecture where we adopted an activity tracker as our sensing device to reduce cumbersome for daily use. Among the total of 17 features extracted from activity tracker, five features from sleep data and six features from heart rate data were proposed to develop the stress recognition model. In the evaluation of our system, we achieved the accuracy as high as 81.70% on the cross validation and 78.95% when tested on the test set. Despite that this is a preliminary result, it has shown that it is possible to use the IoT architecture along with the activity tracker to accurately recognize stress and help improve one’s wellbeing.
KW - Activity tracker
KW - Digital healthcare
KW - Internet of Things
KW - Stress recognition
KW - Wearable sensor
UR - http://www.scopus.com/inward/record.url?scp=85048152396&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85048152396&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-92028-3_2
DO - 10.1007/978-3-319-92028-3_2
M3 - Conference contribution
AN - SCOPUS:85048152396
SN - 9783319920276
T3 - Smart Innovation, Systems and Technologies
SP - 11
EP - 21
BT - Intelligent Decision Technologies 2018 - Proceedings of the 10th KES International Conference on Intelligent Decision Technologies KES-IDT 2018
A2 - Jain, Lakhmi C.
A2 - Czarnowski, Ireneusz
A2 - Howlett, Robert J.
A2 - Vlacic, Ljubo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 10th International KES Conference on Intelligent Decision Technologies, KES-IDT 2018
Y2 - 20 June 2018 through 22 June 2018
ER -