Skip to main content

Table 2 Performance of the family classification model using default parameters of Weka

From: Fangorn Forest (F2): a machine learning approach to classify genes and genera in the family Geminiviridae

Type of evaluation

ML algorithm

Weighted average among all classes

Accuracy

Precision

Recall

F-Measure

MCC

AUC

Using a test set

MLP

0.9444

0.946

0.944

0.944

0.891

0.969

SMO

0.8107

0.815

0.811

0.810

0.625

0.810

RF

0.9542

0.955

0.954

0.954

0.909

0.988

10-fold cross validation

MLP

0.9369

0.937

0.937

0.937

0.871

0.972

SMO

0.8568

0.861

0.857

0.855

0.709

0.844

RF

0.9601

0.960

0.960

0.960

0.919

0.992

Leave-one-out

MLP

0.944

0.944

0.944

0.944

0.886

0.975

SMO

85.597

0.860

0.856

0.854

0.707

0.843

RF

96.228

0.963

0.962

0.962

0.923

0.992

Mean performance

MLP

0.9420

0.9430

0.9417

0.9417

0.8843

0.9700

SMO

0.8411

0.8433

0.8413

0.8339

0.6803

0.8323

RF

0.9588

0.9533

0.9586

0.9586

0.917

0.9906