Table 1 Performance indices for 5-fold cross validation using different ML classifiers.

From: Structure based drug design and machine learning approaches for identifying natural inhibitors against the human αβIII tubulin isotype

Model

Accuracy

AUC

Recall

Precision

F1

Kappa

MCC

TT (Sec)

AdaBoost Classifier (ada)

0.9985

0.9999

0.9981

0.9990

0.9986

0.9970

0.9971

0.0940

Extreme Gradient Boosting (xgboost)

0.9971

0.9999

0.9961

0.9981

0.9971

0.9941

0.9941

0.0940

Light Gradient Boosting Machine (lightgbm)

0.9980

0.9999

0.9981

0.9980

0.9981

0.9961

0.9961

0.2060

CatBoost Classifier (catboost)

0.9971

0.9999

0.9964

0.9981

0.9972

0.9941

0.9941

0.2520

Gradient Boosting Classifier (gbc)

0.9946

0.9998

0.9940

0.9951

0.9946

0.9892

0.9892

0.1100

Logistic Regression (LR)

0.9892

0.9996

0.9934

0.9861

0.9898

0.9782

0.9783

0.7420

Decision Tree Classifier (DT)

0.9941

0.9939

0.9951

0.9934

0.9942

0.9882

0.9882

0.0840

Random Forest Classifier (RF)

0.9588

0.9894

0.9639

0.9578

0.9607

0.9170

0.9173

0.1300

Linear Discriminant Analysis (LDA)

0.9101

0.9676

0.9471

0.8869

0.9159

0.8189

0.8210

0.0840

Extra Trees Classifier (ET)

0.8689

0.9388

0.8575

0.8873

0.8713

0.7376

0.7395

0.1720

K Neighbors Classifier (KNN)

0.7549

0.8225

0.7802

0.7547

0.7669

0.5077

0.5084

0.5020

Naive Bayes (NB)

0.7127

0.7934

0.6182

0.7837

0.6893

0.4269

0.4393

0.3700

Quadratic Discriminant Analysis (QDA)

0.5511

0.5522

0.3664

0.6218

0.4521

0.1045

0.1201

0.1020

Dummy Classifier

0.5191

0.5000

1.0000

0.5191

0.6829

0.0000

0.0000

0.0780

SVM - Linear Kernel

0.9352

0.0000

0.9615

0.9266

0.9398

0.8706

0.8793

0.0820

Ridge Classifier

0.9298

0.0000

0.9646

0.9064

0.9345

0.8584

0.8605

0.0760

Back to article page