A note on diffusion limits for stochastic gradient descent
Lanconelli, Alberto, Lauria, Christopher S. A.
–arXiv.org Artificial Intelligence
In the machine learning literature stochastic gradient descent has recently been widely discussed for its purported implicit regularization properties. Much of the theory, that attempts to clarify the role of noise in stochastic gradient algorithms, has widely approximated stochastic gradient descent by a stochastic differential equation with Gaussian noise. We provide a novel rigorous theoretical justification for this practice that showcases how the Gaussianity of the noise arises naturally.
arXiv.org Artificial Intelligence
Oct-20-2022
- Country:
- North America > United States
- New York (0.04)
- California > Los Angeles County
- Los Angeles (0.14)
- Europe
- United Kingdom > England
- Cambridgeshire > Cambridge (0.04)
- Italy > Emilia-Romagna
- Metropolitan City of Bologna > Bologna (0.05)
- United Kingdom > England
- North America > United States
- Genre:
- Research Report (0.50)
- Industry:
- Education (0.49)