Clustering on Matched Synthetic Benchmark N=200 D=100, 90% irrelevant dimensions (Small-n)
3Number of Clusters (K)K-means
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| K-means2026.04 | 3 | 90.9 | 88.5 | — | |
| GMM2026.04 | 3 | 99.2 | 98.7 | — | |
| SPKM2026.04 | 3 | 99.3 | 98.9 | 18.4 | |
| DIVI-Infosplit interval Tsplit=1202026.04 | 3 | 99 | 98.5 | 77.2 | |
| DIVI-NonInfosplit interval Tsplit=1202026.04 | 3 | 86.8 | 87.8 | 18.2 | |
| DIVI-Randomsplit interval Tsplit=1202026.04 | 3 | 70 | 72.8 | 18.6 |