The role of regularization in overparameterized neural networks
2020 59th IEEE Conference on Decision and Control (CDC), 2020•ieeexplore.ieee.org
In this paper, we consider gradient descent on a regularized loss function for training an
overparametrized neural network. We model the algorithm as an ODE and show how
overparameterization and regularization work together to provide the right tradeoff between
training and generalization errors.
overparametrized neural network. We model the algorithm as an ODE and show how
overparameterization and regularization work together to provide the right tradeoff between
training and generalization errors.
In this paper, we consider gradient descent on a regularized loss function for training an overparametrized neural network. We model the algorithm as an ODE and show how overparameterization and regularization work together to provide the right tradeoff between training and generalization errors.
ieeexplore.ieee.org
以上显示的是最相近的搜索结果。 查看全部搜索结果