Skip to main content

Table 8 Classification results of algorithms using leave drug combinations out on HT29

From: EDST: a decision stump based ensemble algorithm for synergistic drug combination prediction

HT29

AUC

F1 score

F1 score*

Recall

Precision

Rank

Rank*

Additive

Antagonism

Synergy

Additive

Antagonism

Synergy

SVM

0.5963

0.3566

0.1974

0.8117

0.1560

0.1381

0.6231

0.2758

0.3253

5.8

5.8

KNN

0.5574

0.3429

0.1641

0.8660

0.1154

0.1147

0.6306

0.3209

0.2517

6.3

8.0

XGB

0.6046

0.3559

0.1864

0.8737

0.1145

0.1378

0.6258

0.3090

0.4061

4.7

5.8

MLP

0.5989

0.3448

0.1826

0.8462

0.1190

0.1418

0.6220

0.3028

0.3072

6.1

6.3

DT

0.5067

0.3710

0.2527

0.6153

0.2574

0.2611

0.6238

0.2332

0.2597

6.1

6.0

GDBT

0.6073

0.3655

0.2009

0.8641

0.1424

0.1435

0.6374

0.3020

0.3738

3.7

4.5

RF

0.6121

0.3451

0.1884

0.8180

0.1137

0.1763

0.6279

0.2123

0.3260

5.7

6.1

Single-layer Network

0.5284

0.2031

0.2482

0.3675

0.1728

0.4745

0.3439

0.1454

0.2572

8.1

6.0

DeepSynergy

0.5118

0.2526

0.0576

0.8470

0.1346

0.0000

0.6155

0.1007

0.0000

9.1

10.1

MatchMaker

0.5777

0.3124

0.3177

0.3759

0.5420

0.3039

0.5733

0.2605

0.2483

7.1

5.0

EDST

0.6129

0.4159

0.3283

0.5945

0.3464

0.3453

0.6499

0.3504

0.2746

2.8

2.1

  1. The symbol [bold] indicates the highest value in the same evaluation indicator