An Iterative Dimensionality-Scaling System for Real-Time Health Monitoring Applications

Haik Kalantarian, Majid Sarrafzadeh, Shibo Zhang, Nabil Alshurafa

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

Abstract

Wearable health-monitoring systems must achieve a balance between the often opposing goals of hardware overhead and classification accuracy. Prior works have presented various approaches to dynamically scale the accuracy of these systems as a function of available resources. In this paper, we present a framework which retroactively improves the accuracy of prior estimates when resources become available, using a novel global cost minimization function. We benchmark our algorithm on an audio-based nutrition monitoring dataset. Results confirm the efficacy of our technique.

Original languageEnglish (US)
Title of host publicationProceedings - 2016 IEEE International Conference on Healthcare Informatics, ICHI 2016
EditorsWai-Tat Fu, Kai Zheng, Larry Hodges, Gregor Stiglic, Ann Blandford
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages488-494
Number of pages7
ISBN (Electronic)9781509061174
DOIs
StatePublished - Dec 6 2016
Event2016 IEEE International Conference on Healthcare Informatics, ICHI 2016 - Chicago, United States
Duration: Oct 4 2016Oct 7 2016

Publication series

NameProceedings - 2016 IEEE International Conference on Healthcare Informatics, ICHI 2016

Other

Other2016 IEEE International Conference on Healthcare Informatics, ICHI 2016
Country/TerritoryUnited States
CityChicago
Period10/4/1610/7/16

ASJC Scopus subject areas

  • Health Informatics
  • Health(social science)
  • Computer Networks and Communications
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'An Iterative Dimensionality-Scaling System for Real-Time Health Monitoring Applications'. Together they form a unique fingerprint.

Cite this