Active transfer learning for activity recognition

Tom Diethe, Niall Twomey, Peter Flach

Research output: Contribution to conferenceConference Paperpeer-review

177 Downloads (Pure)

Abstract

We examine activity recognition from accelerometers, which provides at least two major challenges for machine learning. Firstly, the deployment context is likely to differ from the learning context. Secondly, accurate labelling of training data is time-consuming and error-prone. This calls for a combination of active and transfer learning. We derive a hierarchical Bayesian model that is a natural fit to such problems, and provide empirical validation on synthetic and publicly available datasets. The results show that by combining active and transfer learning, we can achieve faster learning with fewer labels on a target domain than by either alone.
Original languageEnglish
Number of pages6
Publication statusPublished - 2016
EventEuropean Symposium on Artificial Neural Networks - Belgium, Bruges, United Kingdom
Duration: 27 Apr 201629 Apr 2016

Conference

ConferenceEuropean Symposium on Artificial Neural Networks
CountryUnited Kingdom
CityBruges
Period27/04/1629/04/16

Structured keywords

  • Jean Golding
  • SPHERE

Keywords

  • active learning
  • transfer learning
  • smart home
  • accelerometers

Fingerprint

Dive into the research topics of 'Active transfer learning for activity recognition'. Together they form a unique fingerprint.

Cite this