Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

How Many Factors Influence Minima in SGD?

Sep 24, 2020
Victor Luo, Yazhen Wang

Stochastic gradient descent (SGD) is often applied to train Deep Neural Networks (DNNs), and research efforts have been devoted to investigate the convergent dynamics of SGD and minima found by SGD. The influencing factors identified in the literature include learning rate, batch size, Hessian, and gradient covariance, and stochastic differential equations are used to model SGD and establish the relationships among these factors for characterizing minima found by SGD. It has been found that the ratio of batch size to learning rate is a main factor in highlighting the underlying SGD dynamics; however, the influence of other important factors such as the Hessian and gradient covariance is not entirely agreed upon. This paper describes the factors and relationships in the recent literature and presents numerical findings on the relationships. In particular, it confirms the four-factor and general relationship results obtained in Wang (2019), while the three-factor and associated relationship results found in Jastrz\c{e}bski et al. (2018) may not hold beyond the considered special case.

Share this with someone who'll enjoy it:

   Access Paper Source

Share this with someone who'll enjoy it: