Table 1 Machine learning models of arousal detection

[3]

Acrophobia

GSR, HR, BVR

SVM, RF, k-NN

SVM-42.6%, k-NN-89.5%, RF-99%

Unimodal

[62]

Drug addiction

HRV

PCA, k-Means++

.

Unimodal

[63]

Spider phobia

Clinical characteristics

RF, Permutation Test

\(*p < 0.05\); \(**p < 0.01\); \(***p < 0.001\)

Unimodal

[64]

Spider phobia

fMRI, genetic data

SVM, GPC

.

Unimodal

[65]

PSA

.

.

.

Unimodal

[66]

Anxiety disorder

EEG

SVM

Healthy subjects-97.70 ± 3.32%, Anxious subjects-92.29 ± 4.44%

Unimodal

[67]

Stress

EEG

k-NN with GA-based feature selection

k-NN 71.76%

Unimodal

[68]

Emotion recognition

EEG

SVM, RF

RF-74.0%, SVM-57.2%

Unimodal

[69]

Major depressive disorder

EEG

k-NN, SVM, CNN

CNN-94.13%, SVM-88.22%, k-NN-83.15%

Unimodal

[70]

Stress

EEG

SVM, NB

SVM-90%, NB-81.7%

Unimodal

[56]

Human affective state

EEG

LR, SVM, NB

Multimodal

[57]

Metal stress

EEG

LR, SVM, NB

85.71%

Unimodal

[71]

Construction worker stress

EEG

k-NN, GDA, SVM

k-NN - 65.80%, QSVM-69.62%, GSVM-80.32%

Unimodal