[Retracted] CCPIN: Classification and Combine Parallel Interaction Network for CTR Prediction

Table 1

Performance evaluation of different models on the dataset.

Model	Criteo		MovieLens-1M
	AUC	Logloss	AUC	Logloss
Wide and Deep	0.8029	0.4456	0.7991	0.4562
DeepFM	0.8059	0.4454	0.7995	0.4451
DCN	0.8042	0.4463	0.7993	0.4459
XdeepFM	0.8069	0.4445	0.8011	0.4447
XCrossNet	0.8098	0.4370	0.8066	0.4439
CCPIN	0.8182	0.4354	0.8084	0.4445

In the above table, in order to see clearly, the metrics of the CCPIN model are bloated. The AUC is 0.8182 and 0.8084 on the Criteo dataset and MovieLens-1M dataset, respectively, and the Logloss is 0.43540 and 0.4445, respectively.