Widrow-Hoff LMS Adaline Demonstrator for Schools and Colleges

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The Widrow-Hoff LMS (or ‘Adaline’) algorithm developed
originally in 1960 is fundamental to the operation of countless
signal processing machine learning systems in use even
today. Bernard Widrow and Ted Hoff famously developed
an Adaline machine demonstrator using basic analog off
the shelf components to show how a ‘perceptron’ could
be trained manually [1]. This paper details the design and
development of a fully digital Adaline Least-Mean-Square
algorithm demonstrator. The simplistic design presented here
is completely open-source with all code, bill of materials
and 3D casing models available at this Github Repository for
download and reproduction. The demonstrator enables quick
visualisation of the training and testing of a simple perceptron
algorithm running on the inexpensive Arduino platform. The
total costs of the device is estimated to be less than $60 and
could be used in classrooms and colleges the world-over to
demonstrate the seminal work of Widrow and Hoff [2] to a
wide audience.
Original languageEnglish
Title of host publicationInternational Conference on Acoustics, Speech and Signal Processing (ICASSP)
Place of PublicationSouth Korea
PublisherIEEE
Publication statusAccepted/In press - 15 Apr 2024

Keywords

  • Widrow-Hoff
  • Adaline
  • LMS
  • Educational Demonstrator

Cite this