Generalization bounds for deep convolutional neural networks

Philip M. Long, Hanie Sedghi

Keywords: cnn, deep learning theory, generalization

Abstract: We prove bounds on the generalization error of convolutional networks. The bounds are in terms of the training loss, the number of parameters, the Lipschitz constant of the loss and the distance from the weights to the initial weights. They are independent of the number of pixels in the input, and the height and width of hidden feature maps. We present experiments using CIFAR-10 with varying hyperparameters of a deep convolutional network, comparing our bounds with practical generalization gaps.

Similar Papers

Understanding Generalization in Recurrent Neural Networks
Zhuozhuo Tu, Fengxiang He, Dacheng Tao,
Towards Stable and Efficient Training of Verifiably Robust Neural Networks
Huan Zhang, Hongge Chen, Chaowei Xiao, Sven Gowal, Robert Stanforth, Bo Li, Duane Boning, Cho-Jui Hsieh,