SchizoGoogLeNet: The GoogLeNet-Based Deep Feature Extraction Design for Automatic Detection of Schizophrenia

Table 4

Overall performances of the proposed methods for the unbalanced dataset.

Classifier	Sensitivity (%)	Specificity (%)	Accuracy (%)	Positive predictive value (%)	F1 score (%)
GoogLeNet	93.81	97.02	95.09	97.95	95.83
SVM	99.02	98.58	98.84	99.06	99.04
KNN	98.30	97.60	98.02	98.42	98.36
DT	98.55	97.62	98.18	98.43	98.49
LDA	99.13	98.18	98.75	98.80	98.97

Bold values represent highest performance.