Implicit Regularization in Hierarchical Tensor Factorization and Deep Convolutional Neural Networks
Noam Razin, Asaf Maman, Nadav Cohen
arXiv preprint arXiv:2201.11729, 2022
In the pursuit of explaining implicit regularization in deep learning, prominent focus was given to matrix and tensor factorizations, which correspond to simplified neural networks. It was shown... read more
Implicit Regularization in Tensor Factorization
Noam Razin*, Asaf Maman*, Nadav Cohen
arXiv preprint arXiv:2102.09972, 2021
Implicit regularization in deep learning is perceived as a tendency of gradient-based optimization to fit training data with predictors of minimal “complexity.” The fact that only some types of ... read more