Research Article
A Buffer Overflow Prediction Approach Based on Software Metrics and Machine Learning
Table 8
Decision tree algorithm-specific operation results.
| ā | Before feature selection (%) | After feature selection (%) | Precision | Recall | F1 | Precision | Recall | F1 |
| Good | 95.85 | 93.04 | 94.42 | 96.11 | 93.04 | 94.55 |
| Good Sink | 47.15 | 53.1 | 49.95 | 47.15 | 53.1 | 49.95 |
| Good Source | 0 | 0 | 0 | 0 | 0 | 0 |
| Good Auxiliary | 98.56 | 98.13 | 98.35 | 98.56 | 98.13 | 98.34 |
| Bad | 77.27 | 73.99 | 75.6 | 77.27 | 73.99 | 75.6 |
| Bad Sink | 0 | 0 | 0 | 0 | 0 | 0 |
| Bad Source | 95.74 | 99.51 | 96.54 | 93.75 | 99.6 | 96.59 |
| Average | 59.22 | 59.69 | 59.27 | 58.98 | 59.69 | 59.33 |
| Accuracy | 87.44 | 87.51 |
| Time | 11.99 | 7.59 |
|
|