Back to Search
Start Over
Non-convergence of stochastic gradient descent in the training of deep neural networks.
- Source :
-
Journal of Complexity . Jun2021, Vol. 64, pN.PAG-N.PAG. 1p. - Publication Year :
- 2021
-
Abstract
- Deep neural networks have successfully been trained in various application areas with stochastic gradient descent. However, there exists no rigorous mathematical explanation why this works so well. The training of neural networks with stochastic gradient descent has four different discretization parameters: (i) the network architecture; (ii) the amount of training data; (iii) the number of gradient steps; and (iv) the number of randomly initialized gradient trajectories. While it can be shown that the approximation error converges to zero if all four parameters are sent to infinity in the right order, we demonstrate in this paper that stochastic gradient descent fails to converge for ReLU networks if their depth is much larger than their width and the number of random initializations does not increase to infinity fast enough. [ABSTRACT FROM AUTHOR]
- Subjects :
- *RANDOM numbers
*APPROXIMATION error
*GENEALOGY
*INFINITY (Mathematics)
Subjects
Details
- Language :
- English
- ISSN :
- 0885064X
- Volume :
- 64
- Database :
- Academic Search Index
- Journal :
- Journal of Complexity
- Publication Type :
- Academic Journal
- Accession number :
- 149369980
- Full Text :
- https://doi.org/10.1016/j.jco.2020.101540