Research Article
Software Defect Prediction Based on Hybrid Swarm Intelligence and Deep Learning
Table 15
SF1 experimental results.
| | Precision | Recall rate | F-measure | AUC |
| Grid | 0.69 | 0.26 | 0.38 | 0.62 | Grid + self-encoding | 0.71 | 0.29 | 0.40 | 0.64 | Random | 0.71 | 0.28 | 0.40 | 0.63 | Random + self-encoding | 0.71 | 0.34 | 0.46 | 0.66 | PSO | 0.67 | 0.36 | 0.45 | 0.66 | PSO + self-encoding | 0.70 | 0.41 | 0.51 | 0.69 | WPA | 0.72 | 0.42 | 0.52 | 0.69 | WPA + self-encoding | 0.75 | 0.48 | 0.58 | 0.73 | WPA_PSO | 0.73 | 0.49 | 0.59 | 0.73 | WPA_PSO + self-encoding | 0.76 | 0.52 | 0.62 | 0.75 |
|
|