The Development of Intelligent Models for Stress Detection towards Real-world Applications

Hemakom, Apit; Atiwiwat, Danita; Sanguantrakul, Jongsook; Israsena, Pasin

doi:10.1109/jcsse54890.2022.9836256

Cited by 1 publication

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RF-Max, RF-Log2, and RF-Sqrt used 100 DTs. It should be noted that this work is different from our previous work [35], as this work uses more data from a larger number of subjects, more features, more classification algorithms, and takes genders into account for building stress detection models, resulting in entirely different stress detection models.…”

Section: Stress Detection and Multilevel Stress Classificationmentioning

confidence: 96%

“…A more sophisticated feature extraction method called multiscale sample entropy was used in [34] to generate features from EEG signals recorded from 28 subjects with no distinction in genders to train a neural network classifier to quantify levels of enjoyment, thoroughly unrelated to stress, in museum pieces with an accuracy of 98%. We recently proposed ML models for stress detection through 131 ECG and EEG features obtained from 20 subjects whereby some of those features were selected using the filter, wrapper, and hybrid methods for model construction [35]. The highest classification accuracy for low-stress detection was 85% using kNN and RF with only 1 feature (the absolute power of the low gamma band recorded from the T3 electrode), and the highest classification accuracy for high-stress detection was 90% using SVM with 108 features.…”

Section: Electroencephalogrammentioning

confidence: 99%

“…The total number of ECG features was 11, and EEG was 136 (147 in total). All the extracted features from all the conditions, including the baseline, can be downloaded from [69] (doi: 10.5061/dryad.kd51c5bbf).…”

Section: Feature Extractionmentioning

confidence: 99%

“…Multilevel stress classification. While our previous work [35] proposed only models for stress detection without taking genders into account, this work proposes new models for multilevel stress classification with genders being considered. Subsets of selected features were used to individually train and validate NB, LR, AdaBoost, kNN, RF, SVM, and RBF-SVM multilevel classifiers that simultaneously distinguish 2 levels of stress (low stress-Class 1, or high stress-Class 2) from the baseline (no stress, Class 0).…”

Section: Stress Detection and Multilevel Stress Classificationmentioning

confidence: 99%

See 3 more Smart Citations

ECG and EEG based detection and multilevel classification of stress using machine learning for specified genders: A preliminary study

Hemakom,

Atiwiwat,

Israsena

2023

PLoS ONE

View full text Add to dashboard Cite

Mental health, especially stress, plays a crucial role in the quality of life. During different phases (luteal and follicular phases) of the menstrual cycle, women may exhibit different responses to stress from men. This, therefore, may have an impact on the stress detection and classification accuracy of machine learning models if genders are not taken into account. However, this has never been investigated before. In addition, only a handful of stress detection devices are scientifically validated. To this end, this work proposes stress detection and multilevel stress classification models for unspecified and specified genders through ECG and EEG signals. Models for stress detection are achieved through developing and evaluating multiple individual classifiers. On the other hand, the stacking technique is employed to obtain models for multilevel stress classification. ECG and EEG features extracted from 40 subjects (21 females and 19 males) were used to train and validate the models. In the low&high combined stress conditions, RBF-SVM and kNN yielded the highest average classification accuracy for females (79.81%) and males (73.77%), respectively. Combining ECG and EEG, the average classification accuracy increased to at least 87.58% (male, high stress) and up to 92.70% (female, high stress). For multilevel stress classification from ECG and EEG, the accuracy for females was 62.60% and for males was 71.57%. This study shows that the difference in genders influences the classification performance for both the detection and multilevel classification of stress. The developed models can be used for both personal (through ECG) and clinical (through ECG and EEG) stress monitoring, with and without taking genders into account.

show abstract

Section: Stress Detection and Multilevel Stress Classificationmentioning

confidence: 96%

Section: Electroencephalogrammentioning

confidence: 99%

Section: Feature Extractionmentioning

confidence: 99%

Section: Stress Detection and Multilevel Stress Classificationmentioning

confidence: 99%

See 2 more Smart Citations