Enabling Training of Neural Networks on Noisy Hardware

Enabling Training of Neural Networks on Noisy Hardware

Deep neural networks (DNNs) are typically trained using the conventional stochastic gradient descent (SGD) algorithm. However, SGD performs poorly when applied to train networks on non-ideal analog hardware composed of resistive device arrays with non-symmetric conductance modulation characteristics...

Full description

Bibliographic Details
Main Author: Tayfun Gokmen
Format: Article
Language:English
Published: Frontiers Media S.A. 2021-09-01
Series:Frontiers in Artificial Intelligence
Subjects:
learning algorithms
training algorithms
neural network acceleration
Bayesian neural network
in-memory computing
on-chip learning
Online Access:https://www.frontiersin.org/articles/10.3389/frai.2021.699148/full

Internet

https://www.frontiersin.org/articles/10.3389/frai.2021.699148/full

Similar Items