Implementation of artificial neurons with tunable width via magnetic anisotropy

Volume: 119, Issue: 20, Pages: 204101 - 204101
Published: Nov 15, 2021
Abstract
We report an experimental implementation of width-tunable neurons to train a binary neural network. The angle-dependent magnetic behavior in an oxide thin film highly mimics neurons with width-controllable activation window, providing an opportunity to train the activation functions and weights toward binary values. We apply this feature to train the MNIST dataset using a 684-800-10 fully connected network and achieve a high accuracy of 97.4%,...
Paper Details
Title
Implementation of artificial neurons with tunable width via magnetic anisotropy
Published Date
Nov 15, 2021
Volume
119
Issue
20
Pages
204101 - 204101
Citation AnalysisPro
  • Scinapse’s Top 10 Citation Journals & Affiliations graph reveals the quality and authenticity of citations received by a paper.
  • Discover whether citations have been inflated due to self-citations, or if citations include institutional bias.