Top-N Recommendation on MovieLens 20M
0.448NDCG@100VASP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VASP2021.02 | 0.448 | 41.4 | 55.2 | |
| RecVAE + PIAEncoder architecture=densely connected layers, Personalized Item Alignment=True2025.11 | 0.446 | 41.7 | 55.6 | |
| H+Vamp Gated2021.02 | 0.445 | 41.3 | 55.1 | |
| FLVAE2021.02 | 0.445 | 40.9 | 54.7 | |
| RecVAE2021.02 | 0.442 | 41.4 | 55.3 | |
| RecVAEEncoder architecture=densely connected layers2025.11 | 0.442 | 41.4 | 55.3 | |
| Multi-VAE + PIAEncoder architecture=MLP, Personalized Item Alignment=True2025.11 | 0.437 | 40.8 | 54.6 | |
| RaCTEncoder architecture=MLP2025.11 | 0.434 | 40.3 | 54.3 | |
| Neural EASE2021.02 | 0.431 | 40.3 | 53.2 | |
| LambdaNetcategory=Matrix factorization & Linear regression2025.11 | 0.427 | 39.5 | 53.4 | |
| Mult-VAE2021.02 | 0.426 | 39.5 | 53.7 | |
| PLReccategory=Matrix factorization & Linear regression2025.11 | 0.426 | 39.4 | 52.7 | |
| Multi-VAEEncoder architecture=MLP2025.11 | 0.426 | 39.5 | 53.7 | |
| EASE2021.02 | 0.42 | 39.1 | 52.1 | |
| EASEcategory=Matrix factorization & Linear regression2025.11 | 0.42 | 39.1 | 52.1 | |
| Mult-DAE2021.02 | 0.419 | 38.7 | 52.4 | |
| CDAEEncoder architecture=MLP2025.11 | 0.418 | 39.1 | 52.3 | |
| GRALScategory=Matrix factorization & Linear regression2025.11 | 0.401 | 37.6 | 50.5 | |
| MFcategory=Matrix factorization & Linear regression2025.11 | 0.399 | 36.7 | 49.8 | |
| WMFcategory=Matrix factorization & Linear regression2025.11 | 0.389 | 36.2 | 49.5 | |
| WARPcategory=Matrix factorization & Linear regression2025.11 | 0.348 | 31 | 44.8 | |
| Popularitycategory=Matrix factorization & Linear regression2025.11 | 0.191 | 16.2 | 23.5 |