Table 4 Assessment of test results
Training set from study group | ‘Jena’ | ‘Jena’ | ‘Berlin’ | ‘Berlin’ | ‘Leipzig’ | ‘Leipzig’ |
|---|---|---|---|---|---|---|
Test set from study group | ‘Berlin’ | ‘Leipzig’ | ‘Jena’ | ‘Leipzig’ | ‘Jena’ | ‘Berlin’ |
Number of rules for ‘CG’ | 9 | 9 | 8 | 8 | 0 | 0 |
Number of rules for ‘OA’ | 10 | 10 | 4 | 4 | 4 | 4 |
Number of rules for ‘RA’ | 10 | 10 | 8 | 8 | 4 | 4 |
Sensitivity for RA (%) | 100 | 100 | 92.3 | 100 | 92.3 | 90 |
Sensitivity for OA (%) | 40 | 100 | 90 | 83.3 | 100 | 100 |
Specificity for RA (%) | 100 | 100 | 80 | 83.3 | 100 | 100 |
Specificity for OA (%) | 100 | 100 | 91.3 | 100 | 92.3 | 90 |
Overall specificity (RA/OA) (%) | 100 | n.a. | 60 | n.a. | n.a. | n.a. |
Accuracy (%) | 80 | 100 | 81.8 | 93.8 | 95.6 | 95 |
Test samples | 30 | 16 | 33 | 16 | 23 | 20 |
Hits for CG | 10 | 0 | 6 | 0 | 0 | 0 |
Hits for OA | 4 | 6 | 9 | 5 | 10 | 10 |
Hits for RA | 10 | 10 | 12 | 10 | 12 | 9 |
Hits total | 24 | 16 | 27 | 15 | 22 | 19 |
Errors for CG | 0 | 0 | 4 | 0 | 0 | 0 |
Errors for OA | 6 | 0 | 1 | 1 | 0 | 0 |
Errors for RA | 0 | 0 | 1 | 0 | 1 | 1 |
Errors total | 6 | 0 | 6 | 1 | 1 | 1 |