Manifold Mixup: Better Representations by Interpolating Hidden States

    arXiv: Machine Learning, 2018.

    Cited by: 19|Bibtex|Views56|

    Abstract:

    Deep neural networks excel at learning the training data, but often provide incorrect and confident predictions when evaluated on slightly different test examples. This includes distribution shifts, outliers, and adversarial examples. To address these issues, we propose manifoldmixup{}, a simple regularizer that encourages neural networks...More

    Code:

    Data:

    Your rating :
    0

     

    Tags
    Comments