Depth with Nonlinearity Creates No Bad Local Minima in ResNetsEIWOS

    Kenji Kawaguchi
    Kenji Kawaguchi
    Cited by: 2|Bibtex|32|

    Neural Networks, Volume abs/1810.090382019,

    Keywords:
    Deep learningLocal minimaNon-convex optimizationResidual neural network

    Abstract:

    In this paper, we prove that depth with nonlinearity creates no bad local minima in a type of arbitrarily deep ResNets with arbitrary nonlinear activation functions, in the sense that the values of all local minima are no worse than the global minimum value of corresponding classical machine-learning models, and are guaranteed to further ...More
    Your rating :
    0

     

    Tags
    Comments