Regularized Gradient Clipping Provably Trains Wide and Deep Neural Networks

arXiv:2404.08624v1 Announce Type: new
Abstract: In this work, we instantiate a regularized form of the gradient clipping algorithm and prove that it can converge to the global minima of deep neural network loss functions provided that the net is of sufficient width. We present empirical evidence that our theoretically founded regularized gradient clipping algorithm is also competitive with the state-of-the-art deep-learning heuristics. Hence the algorithm presented here constitutes a new approach to rigorous deep learning.
The modification we do to …

abstract algorithm arxiv converge cs.lg deep neural network evidence form functions global gradient loss math.oc network networks neural network neural networks prove trains type work