Accuracy and timeliness in ML based activity recognition

Robert J. Ross, John Kelleher

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

4 Citations (Scopus)

Abstract

While recent Machine Learning (ML) based techniques for activity recognition show great promise, there remain a number of questions with respect to the relative merits of these techniques. To provide a better understanding of the relative strengths of contemporary Activity Recognition methods, in this paper we present a comparative analysis of Hidden Markov Model, Bayesian, and Support Vector Machine based human activity recognition models. The study builds on both pre-existing and newly annotated data which includes interleaved activities. Results demonstrate that while Support Vector Machine based techniques perform well for all data sets considered, simple representations of sensor histories regularly outperform more complex count based models.

Original languageEnglish
Title of host publicationPlan, Activity, and Intent Recognition - Papers from the 2013 AAAI Workshop, Technical Report
PublisherAI Access Foundation
Pages39-46
Number of pages8
ISBN (Print)9781577356240
DOIs
Publication statusPublished - 2013
Event2013 AAAI Workshop - Bellevue, WA, United States
Duration: 15 Jul 201315 Jul 2013

Publication series

NameAAAI Workshop - Technical Report
VolumeWS-13-13

Conference

Conference2013 AAAI Workshop
Country/TerritoryUnited States
CityBellevue, WA
Period15/07/1315/07/13

Keywords

  • Machine Learning
  • activity recognition
  • Hidden Markov Model
  • Bayesian
  • Support Vector Machine
  • sensor histories
  • interleaved activities

Fingerprint

Dive into the research topics of 'Accuracy and timeliness in ML based activity recognition'. Together they form a unique fingerprint.

Cite this