Multidimensional Projection on MNIST, FashionMNIST, HAR, Reuters, USPS Aggregated Means
0.94Trustworthinesst-SNE
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| t-SNE2023.06 | 0.94 | 0.968 | 0.436 | 25.951 | 91.4 | 0.754 | |
| UMAP2023.06 | 0.91 | 0.96 | 0.424 | 1.389 | 87.4 | 0.741 | |
| NNP[t-SNE]2023.06 | 0.903 | 0.967 | 0.441 | 0.872 | 85.9 | 0.742 | |
| AE2023.06 | 0.878 | 0.917 | 0.342 | 1.236 | 79.3 | 0.622 | |
| SSNPLabeling Strategy=KM (K-Means)2023.06 | 0.862 | 0.928 | 0.451 | 0.334 | 77.7 | 0.62 | |
| SSNPLabeling Strategy=AG (Agglomerative)2023.06 | 0.849 | 0.922 | 0.461 | 0.301 | 81.2 | 0.674 | |
| ShaRPLabeling Strategy=KM (K-Means)2023.06 | 0.832 | 0.897 | 0.426 | 0.782 | 74.7 | 0.658 | |
| Isomap2023.06 | 0.817 | 0.927 | 0.54 | 4.355 | 73.7 | 0.587 | |
| ShaRPLabeling Strategy=AG (Agglomerative)2023.06 | 0.816 | 0.886 | 0.382 | 0.77 | 77.1 | 0.661 | |
| SSNPLabeling Strategy=GT (Ground Truth)2023.06 | 0.797 | 0.902 | 0.454 | 0.5 | 97.7 | 0.93 | |
| ShaRPLabeling Strategy=GT (Ground Truth)2023.06 | 0.755 | 0.864 | 0.343 | 0.783 | 93.9 | 0.89 |