|
Results of jackknife test |
||||||
| Dataset |
Algorithm |
Rate of correct prediction for each class |
Overall rate of accuracy |
|||
|
|
||||||
| All-α |
All-β |
α/β |
α+β |
|||
|
|
||||||
| 277 domains |
Component coupled |
84.3% |
82.0% |
81.5% |
67.7% |
79.1% |
| Neural network |
68.6% |
85.2% |
86.4% |
56.9% |
74.7% |
|
| SVM |
74.3% |
82.0% |
87.7% |
72.3% |
79.4% |
|
| Rough Sets |
77.1% |
77.0% |
93.8% |
66.2% |
79.4% |
|
|
|
||||||
| 498 domains |
Component coupled |
93.5% |
88.9% |
90.4% |
84.5% |
89.2% |
| Neural network |
86.0% |
96.0% |
88.2% |
86.0% |
89.2% |
|
| SVM |
88.8% |
95.2% |
96.3% |
91.5% |
93.2% |
|
| Rough Sets |
87.9% |
91.3% |
97.1% |
86.0% |
90.8% |
|
Cao et al. BMC Bioinformatics 2006 7:20 doi:10.1186/1471-2105-7-20 |
||||||