Table 5.
Holm statistical tests for precision, recall and F-measure of the algorithm proposed in this thesis and the algorithm proposed by Kong et al. in splitting overlapping cells with α =0 . 1 and the obtained standard scores ( Z 0 )
| Dataset | Metric | z 0 | p-value |
|---|---|---|---|
| Precision | 4.024 | 5.699 ×10−5 | |
| Dataset 1 | Recall | 4.024 | 5.699 ×10−5 |
| F-measure | 3.13 | 0.001 | |
| Precision | 2.683 | 0.007 | |
| Dataset 2 | Recall | 2.459 | 0.001 |
| F-measure | 3.13 | 0.001 | |
| Precision | 4.248 | 2.151 ×10−5 | |
| Dataset 3 | Recall | 3.801 | 1.439 ×10−4 |
| F-measure | 3.801 | 1.439 ×10−4 | |
| Precision | 4.248 | 2.151 ×10−5 | |
| Dataset 4 | Recall | 3.130 | 0.01 |
| F-measure | 2.459 | 0.01 |