Products of Many Large Random Matrices and Gradients in Deep Neural Networks
Sign Up to like & getrecommendations! Published in 2018 at "Communications in Mathematical Physics"
DOI: 10.1007/s00220-019-03624-z
Abstract: We study products of random matrices in the regime where the number of terms and the size of the matrices simultaneously tend to infinity. Our main theorem is that the logarithm of the $$\ell _2$$… read more here.
Keywords: deep neural; many large; neural networks; products many ... See more keywords