Image Classification on Fashion-MNIST uniform noise η=0.4 (7-fold CV)
86.29AccuracyrSDNet
Evaluation results
| Method | Links | |
|---|---|---|
| rSDNetβ=0.1, λ=−12026.03 | 86.29 | |
| rSDNetβ=0.1, λ=−0.82026.03 | 86.23 | |
| GCEq=0.72026.03 | 86.04 | |
| rSDNetβ=0, λ=−0.72026.03 | 85.83 | |
| rSDNetβ=0.05, λ=−0.72026.03 | 85.09 | |
| rSDNetβ=0.05, λ=−0.82026.03 | 84.73 | |
| rSDNetβ=0.1, λ=−0.72026.03 | 84.68 | |
| rSDNetβ=0, λ=−0.82026.03 | 83.65 | |
| rSDNetβ=0.05, λ=−12026.03 | 83.64 | |
| GCEq=0.52026.03 | 78.16 | |
| TCCEparameter=0.32026.03 | 77.94 | |
| rSDNetβ=0, λ=−0.52026.03 | 77.28 | |
| rSDNetβ=0.05, λ=−0.52026.03 | 75.93 | |
| MAE2026.03 | 74.91 | |
| rKLD2026.03 | 72.76 | |
| rSDNetβ=0.1, λ=−0.52026.03 | 72.68 | |
| TCCEparameter=0.22026.03 | 71.37 | |
| rSDNetβ=0.3, λ=−0.82026.03 | 69.41 | |
| rSDNetβ=0.3, λ=−0.72026.03 | 69.03 | |
| rSDNetβ=0.5, λ=−0.82026.03 | 68.89 | |
| rSDNetβ=0.3, λ=−12026.03 | 68.81 | |
| rSDNetβ=0.7, λ=−0.72026.03 | 68.37 | |
| rSDNetβ=0.7, λ=−0.82026.03 | 68.22 | |
| rSDNetβ=0.5, λ=−0.72026.03 | 68.21 | |
| rSDNetβ=0.5, λ=−12026.03 | 68.12 | |
| rSDNetβ=0.5, λ=−0.52026.03 | 67.85 | |
| rSDNetβ=0.7, λ=02026.03 | 67.82 | |
| rSDNetβ=0.7, λ=−12026.03 | 67.77 | |
| rSDNetβ=0.3, λ=−0.52026.03 | 67.76 | |
| SCEα=0.5, β=12026.03 | 67.72 | |
| rSDNetβ=0.7, λ=−0.52026.03 | 67.19 | |
| rSDNetβ=1, λ=02026.03 | 66.86 | |
| TCCEparameter=0.12026.03 | 66.69 | |
| FCLμ=0.502026.03 | 66.44 | |
| rSDNetβ=0.5, λ=02026.03 | 66.33 | |
| rSDNetβ=0.7, λ=0.52026.03 | 66.3 | |
| rSDNetβ=0.3, λ=02026.03 | 65.87 | |
| FCLμ=02026.03 | 65.58 | |
| CCE2026.03 | 65.49 | |
| rSDNetβ=0.1, λ=02026.03 | 65.34 | |
| rSDNetβ=0.5, λ=0.52026.03 | 65.25 | |
| FCLμ=0.252026.03 | 64.51 | |
| FCLμ=0.752026.03 | 10 |