Table 3 A list of common methods to mitigate overfitting
From: Convolutional neural networks: an overview and application in radiology
How to mitigate overfitting | |
|---|---|
More training data | |
Data augmentation | |
Regularization (weight decay, dropout) | |
Batch normalization | |
Reduce architecture complexity |