Click-Through Rate Prediction 벤치마크
Click-Through Rate Prediction on MovieLens 1M
AUC
- 2018-03-09 — RippleNet: AUC 0.921
- 2019-08-12 — KNI: AUC 0.9449
- 2023-08-29 — STEC: AUC 0.9712
| Rank | Model | AUC | Accuracy | Log Loss | Paper | Code | Year |
|---|---|---|---|---|---|---|---|
| 1 | STEC | 0.9712 | – | 0.2016 | STEC: See-Through Transformer-based Encoder for CTR Prediction | 2023 | |
| 2 | KNI | 0.9449 | – | – | An End-to-End Neighborhood-based Interaction Model for Knowledge-enhanced Recommendation | Atomu2014/KNI | 2019 |
| 3 | RippleNet | 0.921 | 84.4 | – | RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems | hwwang55/RippleNet · qibinc/RippleNet-PyTorch · johnnyjana730/MVIN · +6 | 2018 |
| 4 | MKR | 0.917 | 84.3 | – | Multi-Task Feature Learning for Knowledge Graph Enhanced Recommendation | hwwang55/KGCN · hwwang55/MKR · hsientzucheng/MKR.PyTorch | 2019 |
| 5 | DCNv3 | 0.9074 | – | 0.3001 | FCN: Fusing Exponential and Linear Cross Network for Click-Through Rate Prediction | reczoo/FuxiCTR · salmon1802/DCNv3 | 2024 |
| 6 | AutoInt | 0.846 | – | 0.3784 | AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks | shenweichen/DeepCTR · PaddlePaddle/PaddleRec · shenweichen/DeepCTR-Torch · +16 | 2018 |