Table 3 Performance comparison on the IP scenario prediction.

From: Measuring and classifying IP usage scenarios: a continuous neural trees approach

Region

Sichuan

Shandong

Chongqing

Illinois

Metric

Precision

Recall

Auc

Precision

Recall

Auc

Precision

Recall

Auc

Precision

Recall

Auc

SVM

0.8315

0.8735

0.9705

0.9560

0.9289

0.9916

0.9132

0.8956

0.9819

0.8977

0.8183

0.8902

BN

0.6112

0.6818

0.8944

0.8071

0.8074

0.9765

0.7492

0.8039

0.9524

0.4413

0.5613

0.8980

LDA

0.7719

0.7872

0.9553

0.7927

0.8768

0.9810

0.8186

0.8646

0.9717

0.5902

0.8816

0.9554

RF

0.8646

0.8159

0.9771

0.9614

0.9188

0.9936

0.9541

0.8602

0.9904

0.9852

0.5734

0.9250

XgBoost

0.8767

0.8683

0.9732

0.9548

0.9375

0.9947

0.9470

0.9273

0.9913

0.9851

0.8922

0.9708

CatBoost

0.8746

0.7616

0.9226

0.8710

0.9412

0.9805

0.7630

0.7711

0.9464

0.9216

0.7234

0.9003

TabNet

0.8425

0.8143

0.9623

0.9489

0.9275

0.9878

0.9423

0.8952

0.9757

0.8834

0.6128

0.9109

NON

0.7958

0.8483

0.9664

0.9274

0.9172

0.9917

0.9246

0.9152

0.9790

0.9346

0.7415

0.9303

AutoInt

0.8210

0.7704

0.9591

0.9535

0.9358

0.9926

0.9513

0.9006

0.9820

0.9619

0.8896

0.9661

NODE

0.8443

0.8147

0.9762

0.9601

0.9165

0.9904

0.9525

0.8591

0.9901

0.9843

0.5721

0.9239

ODTSR

0.8997

0.8759

0.9822

0.9629

0.9458

0.9954

0.9558

0.9368

0.9922

0.9861

0.9012

0.9876