https://www.selleckchem.com/pr....oducts/gs-441524.htm
We refine a recently proposed class of local entropic loss functions by restricting the smoothening regularization to only a subset of weights. The new loss functions are referred to as partial local entropies. They can adapt to the weight-space anisotropy, thus outperforming their isotropic counterparts. We support the theoretical analysis with experiments on image classification tasks performed with multilayer, fully connected, and convolutional neural networks. The present study suggests how to better exploit the anisotropic nature
Like
Comment
Share
