Skip to main content

Table 3 Performance of support vector machine (SVM) model in identifying single hard seed in six species

From: Non-destructive identification of single hard seed via multispectral imaging analysis in six legume species

Species Accuracy (%) Sensitivity (%) Specificity (%)
Acacia seyal
 Calibration 89.29 100.00 0.00
 CV 89.29 100.00 0.00
 Validation 88.33 100.00 0.00
Galega orientulis
 Calibration 91.79 90.98 92.52
 CV 89.29 87.22 91.16
 Validation 87.50 88.00 87.14
Glycyrrhiza glabra
 Calibration 82.14 57.65 92.82
 CV 67.50 4.71 94.87
 Validation 80.00 46.88 92.05
Medicago sativa
 Calibration 95.00 93.75 96.32
 CV 86.79 83.33 90.44
 Validation 89.17 84.62 92.65
Melilotus officinalis
 Calibration 90.36 73.08 97.03
 CV 88.57 71.79 95.05
 Validation 91.67 76.67 96.67
Thermopsis lanceolata
 Calibration 80.71 20.00 99.07
 CV 80.00 20.00 98.14
 Validation 77.50 7.41 97.85
  1. CV  cross validation