Table 10 Performance of GCNs with different numbers of layers in depth sensitivity Experiments.

Citeseer

single-layer GCN

0.583929

0.822334

0.385114

0.559663

0.504300

0.356296

0.459963

0.814350

0.515353

two-layer GCN

0.484069

0.818214

0.377790

0.599940

0.568800

0.345601

0.456540

0.842465

0.492008

three-layer GCN

0.444321

0.816524

0.376340

0.617072

0.605500

0.336664

0.459849

0.792908

0.515290

Acm

single-layer GCN

0.684481

0.745318

0.620495

0.844300

0.670900

0.692292

0.788906

0.798375

0.518493

two-layer GCN

0.589847

0.752684

0.474369

0.621157

0.524600

0.475670

0.626644

0.598632

0.507636

three-layer GCN

0.572061

0.751419

0.440557

0.606281

0.599700

0.435782

0.595400

0.670572

0.502263

Amap

single-layer GCN

0.377126

0.670459

0.646764

0.685882

0.684800

0.506579

0.607000

0.899968

0.791961

two-layer GCN

0.536609

0.661113

0.641394

0.608366

0.629600

0.486151

0.604802

0.751231

0.251217

three-layer GCN

0.566904

0.659047

0.603202

0.599869

0.621300

0.468950

0.587254

0.692937

0.487811

Cocs

single-layer GCN

0.534169

0.646346

0.528356

0.624284

0.527100

0.508426

0.586879

0.823781

0.681961

two-layer GCN

0.636651

0.614536

0.390344

0.522064

0.392900

0.382378

0.482495

0.544767

0.381217

three-layer GCN

0.479450

0.591030

0.289053

0.442481

0.337100

0.315236

0.411786

0.717388

0.327811