Implementation of quadric perceptron with hardlims activation function in a FPGA for nonlinear pattern classification

Raymundo Cordero Garcia, Walter Issamu Suemitsu, Joao Onofre Pereira Pinto, Andre Muniz Soares

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper deals with the design and implementation of an artificial neural network for pattern classification in the FPGA EP2C20F484C7. A perceptron with quadratic decision boundary is used as nonlinear classification system, but using a hardlims as activation function, instead of a sigmoid function. The training algorithm is similar to the used in conventional perceptron. The elimination of the sigmoid function makes simpler the implementation of quadratic perceptrons. As the mentioned FPGA does not do neither float-point nor fixed-point multiplications, the synaptic weights were normalized to integers. The proposed quadratic perceptron is tested in a set of classification problems and compared with multilayer perceptron. Example of experimental implementation of the proposed classification system is shown, including parameters about computational cost.

Original languageEnglish
Title of host publicationProceedings, IECON 2013 - 39th Annual Conference of the IEEE Industrial Electronics Society
Pages2432-2437
Number of pages6
DOIs
StatePublished - 2013
Externally publishedYes
Event39th Annual Conference of the IEEE Industrial Electronics Society, IECON 2013 - Vienna, Austria
Duration: Nov 10 2013Nov 14 2013

Publication series

NameIECON Proceedings (Industrial Electronics Conference)

Conference

Conference39th Annual Conference of the IEEE Industrial Electronics Society, IECON 2013
Country/TerritoryAustria
CityVienna
Period11/10/1311/14/13

Keywords

  • Artificial neural networks
  • FPGA
  • decision boundary
  • multilayer perceptron
  • quadratic perceptron

Fingerprint

Dive into the research topics of 'Implementation of quadric perceptron with hardlims activation function in a FPGA for nonlinear pattern classification'. Together they form a unique fingerprint.

Cite this