Skip to main content

Table 3 Parameters of classifiers and discretizers

From: Graph clustering-based discretization of splitting and merging methods (GraphS and GraphM)

Method

Parameters

Classifier

 

 C4.5

Pruned tree \(=true\), confidence \(=0.25\), minimum example per leaf \(=2\)

 KNN

\(k=3\), distanced function \(=EuclideanDistance\)

Discretizer

 

 ChiMerge

Confidence threshold \(=0.05\)

 EMD

Population size \(=50, ~ M_e=10,000, ~ \alpha =0.7, ~ R_{rate}=0.1, ~ R_{perc}=0.5\)

 FFD

Frequency size \(=30\)

 FUSINTER

\(\alpha =0.975, ~ \lambda =1\)

 HDD

Coefficient \(=0.8\)

 GraphS, GraphM

\(\beta =1.01\)