Binary classification on Synthesized persistence diagrams 100 (test)
86AccuracyPWGK
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PWGKkernel=kG, weight=warc, p=10, SVM kernel=Gaussian2017.06 | 86 | 4.98 | |
| PWGKkernel=kG, weight=warc, p=1, SVM kernel=Gaussian2017.06 | 85.8 | 5.19 | |
| PWGK with Persistence imageMesh size (M)=100, weight=warc, p=5, SVM kernel=Gaussian2017.06 | 85.8 | 4.15 | |
| PWGKkernel=kG, weight=warc, p=5, SVM kernel=Gaussian2017.06 | 85.6 | 5.01 | |
| PWGKkernel=kG, weight=warc, p=10, SVM kernel=Linear2017.06 | 76 | 2.39 | |
| PWGKkernel=kG, weight=warc, p=5, SVM kernel=Linear2017.06 | 75.8 | 2.47 | |
| PWGKkernel=kG, weight=warc, p=1, SVM kernel=Linear2017.06 | 75.7 | 2.31 | |
| PWGK with Persistence imageMesh size (M)=100, weight=warc, p=5, SVM kernel=Linear2017.06 | 75 | 2.2 | |
| PWGKkernel=kL, weight=wpers, SVM kernel=Gaussian2017.06 | 55.7 | 8.68 | |
| PWGKkernel=kG, weight=wone, SVM kernel=Gaussian2017.06 | 55.1 | 8.42 | |
| Persistence imageMesh size (M)=100, weight=wpers, SVM kernel=Linear2017.06 | 54.5 | 8.76 | |
| PWGKkernel=kG, weight=wone, SVM kernel=Linear2017.06 | 53.8 | 4.76 | |
| PSSKSVM kernel=Gaussian2017.06 | 53.6 | 6.69 | |
| PWGKkernel=kL, weight=wone, SVM kernel=Gaussian2017.06 | 53 | 4.89 | |
| PWGKkernel=kG, weight=wpers, SVM kernel=Gaussian2017.06 | 52.3 | 6.6 | |
| Persistence imageMesh size (M)=50, weight=wpers, SVM kernel=Gaussian2017.06 | 52.3 | 7.21 | |
| Persistence imageMesh size (M)=100, weight=wpers, SVM kernel=Gaussian2017.06 | 52.1 | 6.7 | |
| PWGK with Persistence imageMesh size (M)=20, weight=warc, p=5, SVM kernel=Gaussian2017.06 | 52 | 5.65 | |
| PWGKkernel=kL, weight=warc, p=5, SVM kernel=Gaussian2017.06 | 51.8 | 3.52 | |
| PWGK with Persistence imageMesh size (M)=50, weight=warc, p=5, SVM kernel=Gaussian2017.06 | 51.8 | 7.23 | |
| Persistence imageMesh size (M)=20, weight=wpers, SVM kernel=Gaussian2017.06 | 51.7 | 6.86 | |
| Persistence imageMesh size (M)=20, weight=wpers, SVM kernel=Linear2017.06 | 51.1 | 4.38 | |
| PWGKkernel=kL, weight=wpers, SVM kernel=Linear2017.06 | 51 | 6.84 | |
| Persistence landscapeSVM kernel=Linear2017.06 | 50.6 | 5.92 | |
| PWGKkernel=kL, weight=wone, SVM kernel=Linear2017.06 | 50.5 | 6.9 | |
| PSSKSVM kernel=Linear2017.06 | 50.5 | 5.6 | |
| PWGKkernel=kG, weight=wpers, SVM kernel=Linear2017.06 | 49.3 | 2.72 | |
| PWGKkernel=kL, weight=warc, p=5, SVM kernel=Linear2017.06 | 49.3 | 6.92 | |
| PWGK with Persistence imageMesh size (M)=50, weight=warc, p=5, SVM kernel=Linear2017.06 | 49.2 | 5.77 | |
| Persistence imageMesh size (M)=50, weight=wpers, SVM kernel=Linear2017.06 | 49 | 6.14 | |
| PWGK with Persistence imageMesh size (M)=20, weight=warc, p=5, SVM kernel=Linear2017.06 | 48.8 | 3.75 | |
| Persistence landscapeSVM kernel=Gaussian2017.06 | 45.8 | 4.25 |