neural-networkcomputer-visionconv-neural-networkoverfitting-underfitting

How to improve perfomance of CNN and reduce overfitting?

I am working on a small computer vision project and I'm using convolutional nets for classification. I have already used dropout, l1, l2 regularization and data augmentation to reduce overfitting. Are there any other techniques and algorithms for improving model accuracy and reducing overfitting?

Solution

  • there could be a 100 solutions

    1. Use pretrained model (transfer learning).
    2. Try to implement a smaller network.
    3. Bigger data set.
    4. Try different parameters[learning rate, batch size..].
    5. Use grid search for these parameters.
    6. Try data augmentation for your training data set. ...