Table 11. Classification on the combination of fixation report and demographic data sets: The average and standard deviation of evaluation metrics over 10 different data splits.

The best results are bold-faced and the second ones are underlined.

Methods	Metrics
	Precision	Recall	F1-score	ROC-AUC
Logistic Regression	0.573 ± 0.003	0.658 ± 0.002	0.599 ± 0.003	0.713 0.003
Gaussian Naive Bayes	0.724 ± 0.018	0.302 ± 0.001	0.162 ± 0.001	0.689 ± 0.003
Support Vector	0.807 ± 0.003	0.807 ± 0.003	0.802 ± 0.003	0.872 ± 0.002
K-Nearest Neighbour	0.903 ± 0.001	0.903 ± 0.001	0.903 ± 0.001	0.976 ± 0.001
Random Forest	0.911 ± 0.002	0.910 ± 0.002	0.911 ± 0.002	0.981 ± 0.001
Gradient Boosting	0.902 ± 0.002	0.901 ± 0.003	0.901 ± 0.003	0.978 ± 0.001
AdaBoost	0.669 ± 0.007	0.684 ± 0.002	0.642 ± 0.003	0.724 ± 0.002
Multi-Layer Perceptron	0.913 ± 0.002	0.911 ± 0.002	0.912 ± 0.002	0.983 ± 0.000
Convolutional neural networks	0.657 ± 0.097	0.649 ± 0.096	0.641 ± 0.096	0.713 ± 0.111
Fused CNN-MLP	0.685 ± 0.098	0.692 ± 0.094	0.675 ± 0.097	0.767 ± 0.098