Skip to main content

Advertisement

Table 3 Prediction performances of seven classifiers on discriminating PM+ and PM−, evaluated using the fivefold cross validation technique and full data

From: Statistical prediction of immunity to placental malaria based on multi-assay antibody data for malarial antigens

Classifier Overall Accuracy Specificity Sensitivity AUC
LR-Full 0.68 0.51 0.73 0.64
LR-Reduced 0.71 0.66 0.73 0.76
RPART 0.69 0.54 0.74 0.67
RF 0.75 0.30 0.90 0.74
LDA 0.45 0.82 0.33 0.73
QDA 0.49 0.63 0.46 0.71
SVM 0.51 0.42 0.83 0.69
  1. LR-Full Logistic Regression Full model, LR-Reduced Logistic Regression Reduced model, RPART Recursive Partitioning, RF Random Forests, LDA Linear Discriminant Analysis, QDA Quadratic Discriminant Analysis, SVM Support Vector Machine