Table 5 The ranges of hyperparameters used for training ML algorithms
From: Screening ovarian cancer by using risk factors: machine learning assists
Algorithm | Ranges of parameters used as grid-search technique |
|---|---|
ANN | Number of hidden layers [5,20]; learning rate [0.3,1]; validation threshold [20,100] |
KNN | K [3,7] |
J-48 | Number of objects [1,5]; confidence factor [0.15,0.45]; number of fold [2,6] |
RF | MAX_Depth [6,20]; number of randomly chosen features [5,20] |
SVM | c [1,100]; RBF-gamma [0.1,1] |
XG-Boost | Eta [0.3,1]; Gamma [0,2]; max_depth [5,20]; min_child_weight [0,5] |