Rapid and non-destructive identification of Anopheles gambiae and Anopheles arabiensis mosquito species using Raman spectroscopy via machine learning classification models

Table 1 Summary of quality metrics used to assess the performance of the four classification models

Model	Accuracy(%)		Sensitivity (%)		Specificity (%)
Model	Cross-validation set	Test set	Cross-validation set	Test set	Cross-validation set	Test set
Linear Discriminant Analysis (LDA)	86	79	82	68	90	90
Quadratic Discriminant Analysis (QDA)	89	81	84	66	93	96
Logistic Regression (LR)	85	82	84	74	87	90
Quadratic Support Vector Machine (QSVM)	93	93	90	90	95	96

Total number of mosquitoes used for cross-validation = 297 (Anopheles gambiae, n = 151; Anopheles arabiensis, n = 146); Total number of mosquitoes used for model testing = 100 (Anopheles gambiae, n = 50; Anopheles arabiensis, n = 50)

ISSN: 1475-2875