Table 3.
Features of the artificial intelligence (AI; N=19).
| Feature | Values | References | |
| Problem-solving approaches, n (%) | |||
|
|
Classification | 18 (95) | [46-63] |
|
|
Regression | 1 (5) | [50] |
|
|
Clustering | 1 (5) | [64] |
| Number of stress classes, n (%) | |||
|
|
2 | 15 (79) | [46,47,49,51,53-62,64] |
|
|
3 | 5 (26) | [48,50,52,57,61] |
|
|
>3 | 2 (11) | [55,63] |
| AI algorithms, n (%) | |||
|
|
Support vector machine | 14 (74) | [46,48,50-57,59,60,62,63] |
|
|
K-nearest neighbor | 13 (68) | [46,47,50,52-57,61-63] |
|
|
Random forest | 8 (42) | [47,50,51,54,55,61-63] |
|
|
Logistic regression | 6 (32) | [46,48,51,53,62,63] |
|
|
Decision tree | 5 (26) | [50,52,55,58,61] |
|
|
Naïve Bayes | 5 (26) | [46,50,51,58,62] |
|
|
Artificial neural network | 4 (21) | [53,55,62,63] |
|
|
Multilayer perceptron | 4 (21) | [46,48,54,56] |
|
|
Others | <4 | [46-50,52,55,56,60-64] |
| Aim of AI algorithms, n (%) | |||
|
|
Detection | 19 (100) | [46-64] |
| Data set size, mean (SD; range) | 518 (409; 30-1178) | [46,47,52,53,55,61-64] | |
| Data sources, n (%) | |||
|
|
Wearable devices | 19 (100) | [46-64] |
|
|
Self-reported questionnaire | 11 (58) | [47,49,51,53-55,58,60,62,63] |
|
|
Nonwearable devices | 6 (32) | [47,53,54,58,60,61] |
| Data types, n (%) | |||
|
|
Heart rate data | 16 (84) | [46-54,56,58,59,61,63,64] |
|
|
Electrodermal activity data | 10 (53) | [48,50-52,54,59-61,63,64] |
|
|
Activity data | 8 (42) | [47,48,53-55,58,60,61] |
|
|
Sleep data | 5 (26) | [53-55,58,60] |
|
|
Skin temperature | 3 (16) | [48,50,60] |
|
|
Smartphone use data | 3 (16) | [47,54,60] |
|
|
Location | 3 (16) | [47,54,60] |
|
|
Respiratory rate data | 3 (16) | [47,54,61] |
|
|
Others | <3 | [47,53,54,57,58,60,61] |
| Number of features | |||
|
|
Values, mean (SD; range) | 52.2 (115.1; 4-498) | [46-64] |
|
|
Not reported, n (%) | 1 (5) | [55] |
| Stress-induction methods, n (%) | |||
|
|
Arithmetic tasks | 7 (37) | [46,50,51,56,59,61,63] |
|
|
Exams | 4 (21) | [49,50,52,62] |
|
|
Stroop test | 4 (21) | [50,51,56,63] |
|
|
Physical stress | 3 (16) | [50,51,56] |
|
|
Others | <3 | [48,50-52,56,59,61,64] |
|
|
No stress induction | 7 (37) | [47,53-55,57,58,60] |
| Ground truth, n (%) | |||
|
|
Self-reported questionnaire | 11 (58) | [47-49,51,53-55,57,58,60,62] |
|
|
Context | 9 (47) | [48,50-52,56,59,61,63,64] |
|
|
Unsupervised algorithm | 1 (5) | [46] |
| Types of validation, n (%) | |||
|
|
K-fold | 11 (58) | [46-48,50-54,56,57,63] |
|
|
Training-test split | 6 (32) | [46,49,55,57,59,62] |
|
|
Leave-one-out cross-validation | 2 (11) | [51,61] |
|
|
Nested | 1 (5) | [60] |
|
|
Not reported | 1 (5) | [58] |
|
|
Not applicable | 1 (5) | [64] |
| Performance measures, n (%) | |||
|
|
Accuracy | 16 (84) | [46,48-50,52-61,63,64] |
|
|
F1-score | 9 (47) | [46-48,51,55,56,60,61,64] |
|
|
Precision | 7 (37) | [48,51,55,56,58,59,64] |
|
|
Sensitivity | 6 (32) | [48,55,56,58,62,64] |
|
|
κ | 3 (16) | [46,50,51] |
|
|
Others | <4 | [46,50,51,58,62,64] |