Skip to main content

Table 5 Evaluation results of Diabetes dataset using different sampling strategies with three metrics across ten classification algorithms.

From: A particle swarm based hybrid system for imbalanced medical data sampling

Method

Metric

Classifier

  

J48

3NN

NB

RF5

LOG

1NN

7NN

SMO

RF10

RBFNet

R. Avg.

PSO

AUC

0.746

0.761

0.808

0.801

0.827

0.693

0.793

0.740

0.817

0.786

0.777

 

FMeasure

0.660

0.618

0.638

0.662

0.661

0.612

0.651

0.662

0.671

0.639

0.647

 

GMean

0.734

0.698

0.717

0.736

0.734

0.691

0.727

0.738

0.745

0.719

0.724

 

C. Avg.

0.713

0.692

0.721

0.733

0.741

0.665

0.724

0.713

0.744

0.715

0.716

RU

AUC

0.697

0.739

0.801

0.765

0.829

0.665

0.773

0.737

0.791

0.759

0.756

 

FMeasure

0.635

0.603

0.635

0.628

0.665

0.581

0.636

0.657

0.659

0.609

0.631

 

GMean

0.707

0.686

0.714

0.705

0.740

0.660

0.714

0.734

0.734

0.694

0.709

 

C. Avg.

0.680

0.676

0.717

0.699

0.745

0.635

0.708

0.709

0.728

0.687

0.699

RO

AUC

0.709

0.722

0.799

0.774

0.831

0.653

0.774

0.735

0.796

0.797

0.760

 

FMeasure

0.634

0.592

0.628

0.612

0.665

0.549

0.621

0.656

0.616

0.636

0.622

 

GMean

0.713

0.675

0.705

0.696

0.739

0.643

0.699

0.733

0.696

0.716

0.702

 

C. Avg.

0.685

0.663

0.711

0.694

0.745

0.615

0.698

0.708

0.703

0.716

0.695

Cluster

AUC

0.701

0.729

0.801

0.759

0.813

0.608

0.769

0.753

0.788

0.784

0.751

 

FMeasure

0.624

0.603

0.635

0.598

0.684

0.513

0.626

0.680

0.643

0.629

0.624

 

GMean

0.702

0.678

0.711

0.686

0.723

0.615

0.698

0.752

0.721

0.711

0.700

 

C. Avg.

0.676

0.670

0.716

0.681

0.740

0.579

0.698

0.728

0.717

0.708

0.692