Skip to main content

Table 6 Testing results of different features rank (10 Fold CV)

From: Vote-Based Feature Selection Method for Stratigraphic Recognition in Tunnelling Process of Shield Machine

Feature Rank

RF

LDA

QDA

SVM

KNN

Rank1+2

0.926

0.894

0.784

0.871

0.902

Rank2

0.918

0.889

0.778

0.871

0.896

Rank3

0.921

0.808

0.120

0.748

0.774

Rank4

0.902

0.893

0.883

0.752

0.909

Feature Rank

NN

NB

Bag

Ada

GBDT

Rank1+2

0.797

0.860

0.911

0.699

0.933

Rank2

0.293

0.854

0.921

0.705

0.914

Rank3

0.669

0.801

0.812

0.601

0.902

Rank4

0.615

0.856

0.894

0.550

0.875