Table 1.
Methods used in Conformal Prediction Studies for medical applications
| First author | Classification method(s) a | Conformal Prediction method(s) |
|---|---|---|
| Pereira [14, 20] | KNN, Naïve Bayes, and ensemble classifiers | Mondrian predictors and CP with scaling |
| Papadopoulos [21–23] | ANN | Mondrian predictors, LCMCP |
| Alnemer [24] | SVM, DT, KNN, ANN | Non-conformity scoreb |
| Devetyarov [25] | Linear rules | Mondrian predictors |
| Lambrou [26–28] | Rule-based, GA, SVM | Based on the evolved decision rule after prediction |
| Luo [29] | SVM | Dynamic Conformal Prediction |
| Schleif [30] | SNG | GLVQ c |
| Balasubramanian [31] | SVM | Computed with respect to both class levels |
| Bellotti [32] | SVM | Inductive Confidence Machine |
aKNN K-nearest neighbor, SVM support vector machine, DT decision tree, ANN artificial neural networks, GA genetic algorithms, SNG supervised neural gas
bNC score calculated by comparing the distance of the new prediction point to all records in the training set that have the same label to its distance to the rest of the training set
cGLVQ Generalized Learning Vector Quantizer