Flatter, faster: scaling momentum for optimal speedup of SGD.

Aditya Cowsik Tankut Can Paolo Glorioso

Published in: CoRR (2022)

Keyphrases

optimal solution
optimal design
learning rate
real time
least squares
state space
multi agent
multiscale
information retrieval
training set
evolutionary algorithm
dynamic programming
machine learning
neural network
worst case
orders of magnitude
databases
database