Skip to main content

Table 3 Classification model accuracy for cross-validation, validation, and independent test sets. The classification accuracy, i.e. was a mosquito whose actual age was less than 7 days of age or greater than 7 days of age predicted as “young” or “old,” respectively in cross-validation, validation, or ITS1; or the accuracy of predicting a nulliparous mosquito successfully as “young”, a parous mosquito as “old”, or a sporozoite positive mosquito as “old” (ITS2 and ITS3) is presented. All classifications within sets are binary (i.e. young vs old). If accuracy was significant via McNemar’s Chi-square test, the 5–95% confidence interval is presented in the parenthesis. Degree of significance is demarcated

From: Analysis of near infrared spectra for age-grading of wild populations of Anopheles gambiae

Dataset

Accuracy CV

Accuracy V

ITS1 Accuracy

ITS2 Accuracy

ITS3 Accuracy

Dataset 1

     

PLS

0.7913

0.7727 (0.6216–0.8853)**

0.5507

0.5625

0.5128

ObliqueRF

0.8649

0.7955 (0.647–0.902)***

0.5652

0.625 (0.5096–0.7308)*

0.5128

svmLinear

0.8422

0.8636 (0.7265–0.9483)***

0.6232 (0.4983–0.7371)*

0.6232 (0.4983–0.7371) *

0.5128

Dataset 2

     

PLS

0.9165

0.8421 (0.6875–0.9398)***

0.4493

0.6 (0.4844–0.708)*

0.5385

ObliqueRF

0.9354

0.8684 (0.7191–0.9559)***

0.4058

0.55

0.5385

svmLinear

0.9356

0.8947 (0.752–0.9706)***

0.4348

0.6 (0.4844–0.708)*

0.5769

Dataset 3

     

PLS

0.95

0.878 (0.738–0.9592)***

0.5072

0.4625

0.4872

ObliqueRF

0.9687

0.9756 (0.8714–0.9994)***

0.5942

0.55

0.4744

svmLinear

0.9562

0.9756 (0.8714–0.9994)***

0.5217

0.5375

0.4872

Dataset 4

     

PLS

0.895

0.88 (0.7569–0.9547)***

0.4928

0.5

0.5128

ObliqueRF

0.97

0.98 (0.8935–0.9995)***

0.5072

0.525

0.4615

svmLinear

0.945

0.96 (0.8629–0.9951)***

0.5362

0.55

0.4744

Dataset 5

     

PLS

0.7726

0.7073 (0.5965–0.8026)***

0.5942

0.55

0.5385

ObliqueRF

0.8442

0.7805 (0.6754–0.8644)***

0.6667 (0.5429–0.7756)**

0.525

0.4872

svmLinear

0.8232

0.8049 (0.7026–0.8842)***

0.6812 (0.5579–0.7883)**

0.5875

0.5769

Dataset 6

     

PLS

0.7348

0.748 (0.6617–0.8219)***

0.6812 (0.5579–0.7883)**

0.55

0.4872

ObliqueRF

0.8502

0.8537 (0.7786–0.9109)***

0.6232 (0.4983–0.7371)*

0.625 (0.5096–0.7308) *

0.5256

svmLinear

0.8518

0.8374 (0.7601–0.8978)***

0.6957 (0.5731–0.8008)**

0.5625

0.5

  1. *P < 0.05, **P < 0.01, ***P < 0.001
  2. Abbreviations: CV cross-validation, V validation, ITS independent test set, LV latent variables used if applicable