Table 4 Hyperparameter settings and optimization ranges for the integrated model.

From: An integrated graph neural network model for joint software defect prediction and code quality assessment

Hyperparameter name

Value range

Optimal value

Learning Rate

[0.0001, 0.01]

0.001

Batch Size

[16, 128]

64

Dropout Rate

[0.1, 0.5]

0.3

L2 Regularization

[0.0001, 0.01]

0.005

Task Balance Weight α

[0.3, 0.7]

0.5

Task Balance Weight β

[0.3, 0.7]

0.5

Gradient Clipping

[0.5, 2.0]

1.0

Back to article page