Table 7 Precision@k, NDCG@ comparison on MARS Dataset.

From: Adaptive course recommendation using federated learning and graph convolutional networks in IoT-enhanced e-learning

Metrics

Top @K

Proposed

LightGCN–PKA

FGC

Transformer-based FL

PerSVD-Edu

Precision@K (↑)

@5

0.8901

0.8701

0.8610

0.8502

0.8301

@10

0.8852

0.8752

0.8490

0.8553

0.8212

@15

0.8803

0.8603

0.8510

0.8504

0.8350

@20

0.8924

0.8754

0.8710

0.8555

0.8401

NDCG@K (↑)

@5

0.8753

0.8453

0.8231

0.8254

0.8026

@10

0.8802

0.8702

0.8520

0.8403

0.8120

@15

0.8954

0.8754

0.8633

0.8555

0.8104

@20

0.8955

0.8855

0.8750

0.8656

0.8290

Back to article page