Fig 1.
A, B and C; representative images of unresponsive cases, D, E and F; representative images of responsive cases with anthroponotic cutaneous leishmaniasis due to Leishmania tropica from a major focus, Kerman, southeast of Iran.
Table 1.
Baseline features of unresponsive and responsive patients with anthroponotic cutaneous leishmaniasis.
Fig 2.
Structure of the multilayer perceptron with one hidden layer.
Fig 3.
Overall procedures for modeling.
Fig 4.
A, Learning vector quantization neural network ROC plot was prepared for unresponsive and responsive classes; B, Multipass learning vector quantization neural network ROC plot was prepared for unresponsive and responsive classes; C, Multilayer perceptron neural network ROC plot of unresponsive and responsive classes; D, Support vector machine ROC plot of unresponsive and responsive classes; E, K-nearest neighbors ROC plot of unresponsive and responsive classes.
Fig 5.
A, Confusion matrix for learning vector quantization neural network representing the performance of the model; B, Confusion matrix for multipass learning vector quantization neural network representing the performance of the model; C, Confusion matrix for multilayer perceptron neural network representing the performance of the model; D, Confusion matrix for support vector machine representing the performance of the model; E, Confusion matrix for k-nearest neighbors representing the performance of the model.
Table 2.
Classification results for the whole data.
Table 3.
Removing features in order to determine the effectiveness of each one.