- 【Updated on May 12, 2025】 Integration of CiNii Dissertations and CiNii Books into CiNii Research
- Trial version of CiNii Research Automatic Translation feature is available on CiNii Labs
- Suspension and deletion of data provided by Nikkei BP
- Regarding the recording of “Research Data” and “Evidence Data”
Estimating Working Stressor Based on Pulse Wave
Description
In this paper, we estimate workers' stressors from their pulse waves. When human beings are provided stressful stimulus from outside, their pulse wave would change. Since inexpensive commercial sensors of pulse waves are available nowadays, we can expect to detect stresses of workers in daily working environment. We assume four kinds of stressors. We extract ten kinds of features from the workers' pulse wave. We learn features using the random forest, which is a machine learning algorithm good at discrimination with many input variables. It is possible to create a model to identify four types of stress. We conducted an experiment to verify the effectiveness of the model. As a result of the experiment, the average of the F-measure was 0.32. From this result, it is suggested that the stressor could be identified from the pulse waves.
Journal
-
- Proceedings of the 9th International Conference on Machine Learning and Computing
-
Proceedings of the 9th International Conference on Machine Learning and Computing 23-27, 2017-02-24
ACM
