Table 9 Computational efficiency comparison across different methods.
Method | Training time (hours) | Inference time (ms/file) | Memory usage (GB) | GPU utilization (%) |
|---|---|---|---|---|
Random forest | 0.3 | 2.1 | 1.2 | 0 |
Support vector machine | 1.2 | 3.5 | 2.1 | 0 |
Deep neural network | 2.5 | 4.2 | 4.3 | 65 |
LSTM-based model | 3.8 | 5.8 | 6.2 | 72 |
CodeBERT | 4.7 | 7.4 | 9.1 | 81 |
Graph convolutional network | 5.2 | 8.1 | 8.7 | 78 |
Proposed integrated model | 6.8 | 9.3 | 11.4 | 85 |
