Research Article
Software Defect Prediction Based on Hybrid Swarm Intelligence and Deep Learning
Table 12
PC5 experimental results.
| | Precision | Recall rate | F-measure | AUC |
| Grid | 0.21 | 0.06 | 0.09 | 0.53 | Grid + self-encoding | 0.53 | 0.18 | 0.25 | 0.58 | Random | 0.18 | 0.08 | 0.10 | 0.53 | Random + self-encoding | 0.54 | 0.20 | 0.28 | 0.59 | PSO | 0.37 | 0.31 | 0.33 | 0.63 | PSO + self-encoding | 0.54 | 0.40 | 0.44 | 0.68 | WPA | 0.45 | 0.43 | 0.43 | 0.69 | WPA + self-encoding | 0.58 | 0.48 | 0.51 | 0.72 | WPA_PSO | 0.52 | 0.49 | 0.48 | 0.72 | WPA_PSO + self-encoding | 0.57 | 0.54 | 0.54 | 0.75 |
|
|