Clustering Algorithms Evaluation 벤치마크
Clustering Algorithms Evaluation on MNIST
ARI
- 2021-10-04 — AE+GIT: ARI 77.0
| Rank | Model | ARI | F1-score | NMI | Paper | Code | Year |
|---|---|---|---|---|---|---|---|
| 1 | AE+GIT | 77% | 88% | 81% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |
| 2 | GIT | 42% | 59% | 53% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |
| 3 | k-Means++ | 36% | 50% | 45% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |
| 4 | Spectral Clustering | 33% | 41% | 44% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |
| 5 | SpectACI | 17% | 40% | 33% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |
| 6 | QuickShiftPP | 13% | 45% | 45% | Git: Clustering Based on Graph of Intensity Topology | gaozhangyang/DGC · gaozhangyang/git · yangyucheng000/University · +1 | 2021 |