Research Article
Software Defect Prediction Based on Hybrid Swarm Intelligence and Deep Learning
Table 11
LR1 experimental results.
| | | Precision | Recall rate | F-measure | AUC |
| | Grid | 0.51 | 0.32 | 0.36 | 0.64 | | Grid + self-encoding | 0.61 | 0.39 | 0.47 | 0.68 | | Random | 0.50 | 0.30 | 0.31 | 0.63 | | Random + self-encoding | 0.60 | 0.35 | 0.43 | 0.66 | | PSO | 0.57 | 0.46 | 0.50 | 0.71 | | PSO + self-encoding | 0.64 | 0.56 | 0.59 | 0.76 | | WPA | 0.61 | 0.53 | 0.56 | 0.74 | | WPA + self-encoding | 0.64 | 0.57 | 0.60 | 0.76 | | WPA_PSO | 0.62 | 0.58 | 0.59 | 0.77 | | WPA_PSO + self-encoding | 0.62 | 0.64 | 0.62 | 0.79 |
|
|