Implicit Bias of SGD for Diagonal Linear Networks: a Provable Benefit of Stochasticity.

Scott Pesme Loucas Pillaud-Vivien Nicolas Flammarion

Published in: NeurIPS (2021)

Keyphrases

social networks
closed form
heterogeneous networks
data sets
complex systems
network structure
machine learning
feature selection
training set
covariance matrix
network model
stochastic model