Minimum Precision Requirements for Deep Learning with Biomedical Datasets

Charbel Sakr, Naresh Shanbhag

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

Abstract

Deep neural networks (DNNs) are powerful machine learning models but are typically deployed in large computing clusters due to their high computational and parameter complexity. Many biomedical applications require embedded inference on resource-constrained platforms thus causing a challenge when considering the deployment of DNNs. One method to address this challenge is via reduced precision implementations. We use an analytical method to determine suitable minimum precision requirements of DNNs and show its application to the CHB-MIT EEG seizure detection dataset and the Bonn dataset for brain electrical activity recognition. We show that our method leads to 2 × reduction in average precision and 45% complexity reduction compared to the minimum uniform precision assignment. Compared to a conventional 16-b precision assignment, our method leads to 9 x complexity reduction. Furthermore, we study the impact of network topology on precision and accuracy. Once again we find our method to be 2× more efficient that the uniform assignment for all topologies considered.

Original languageEnglish (US)
Title of host publication2018 IEEE Biomedical Circuits and Systems Conference, BioCAS 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538636039
DOIs
StatePublished - Dec 20 2018
Event2018 IEEE Biomedical Circuits and Systems Conference, BioCAS 2018 - Cleveland, United States
Duration: Oct 17 2018Oct 19 2018

Publication series

Name2018 IEEE Biomedical Circuits and Systems Conference, BioCAS 2018 - Proceedings

Other

Other2018 IEEE Biomedical Circuits and Systems Conference, BioCAS 2018
Country/TerritoryUnited States
CityCleveland
Period10/17/1810/19/18

Keywords

  • Accuracy
  • Complexity
  • Deep learning
  • Neural networks
  • Precision

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Health Informatics
  • Instrumentation
  • Signal Processing
  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'Minimum Precision Requirements for Deep Learning with Biomedical Datasets'. Together they form a unique fingerprint.

Cite this