Research Article
Large Margin Graph Embedding-Based Discriminant Dimensionality Reduction
Table 2
The classification performance on the Yale dataset (%).
| | Method | l = 7 (k0 = 3, a = 0.8) | l = 8 (k0 = 5, a = 0.8) | l = 9 (k0 = 7, a = 0.8) |
| | MFA | 62.344.24 (45) | 63.635.61 (50) | 68.235.71 (45) | | DNE | 67.364.17 (21) | 69.235.33 (26) | 73.235.63 (28) | | MNMDP | 71.233.62 (26) | 72.205.39 (21) | 76.235.29 (21) | | DAG-DNE | 77.584.76 (21) | 77.875.56 (18) | 81.175.26 (16) | | LMGE-DDR | 83.173.21 (40) | 83.784.15 (28) | 86.174.55 (22) |
|
|
