Research Article
A Constrained Algorithm Based NMFα for Image Representation
Table 3
The comparison of classification accuracy rates on the Yale Database.
| | Ac() (%) | CNMF | | | | |
| 2 | 81.64 | 85.23 | 85.93 ± 0.65 | 84.18 ± 0.81 | 86.27 ± 2.46 | 3 | 68.48 | 76.91 | 77.14 ± 0.86 | 77.26 ± 0.42 | 79.01 ± 1.09 | 4 | 64.25 | 66.70 | 68.11 ± 0.67 | 66.12 ± 0.16 | 69.33 ± 2.23 | 5 | 65.82 | 67.71 | 69.14 ± 1.21 | 66.54 ± 1.01 | 70.57 ± 2.22 | 6 | 60.12 | 64.26 | 65.62 ± 0.94 | 63.70 ± 0.94 | 66.92 ± 1.41 | 7 | 59.25 | 62.47 | 63.97 ± 0.72 | 59.83 ± 0.38 | 65.23 ± 1.77 | 8 | 54.40 | 59.26 | 60.12 ± 0.31 | 56.76 ± 0.07 | 60.89 ± 1.28 | 9 | 54.85 | 57.30 | 58.08 ± 0.12 | 54.88 ± 0.14 | 58.98 ± 0.62 | 10 | 52.68 | 56.20 | 57.56 ± 0.03 | 53.83 ± 0.01 | 57.56 ± 0.73 |
| Avg. | 62.39 | 65.89 | 67.30 | 64.79 | 68.31 |
|
|