Detail publikace

Behavioral Patterns Identification in Smart-Home Application

VĚCHET, S. HRBÁČEK, J. RŮŽIČKA, M. MAŠEK, P. KREJSA, J.

Anglický název

Behavioral Patterns Identification in Smart-Home Application

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

en

Originální abstrakt

Detection of behavioral patterns in human everyday life is main idea behind self-learning control system used for controling of a thermostat in a smart house. We use a probabilistic naive Bayessian filter in long-term experiments to recognize patterns in human behaviour. Such patterns are used in the control system to automatically adjusting a temperature in experimental room. The main advantage of such method is no needs for manual setup, the control system is naturally self-learned from the user.

Anglický abstrakt

Detection of behavioral patterns in human everyday life is main idea behind self-learning control system used for controling of a thermostat in a smart house. We use a probabilistic naive Bayessian filter in long-term experiments to recognize patterns in human behaviour. Such patterns are used in the control system to automatically adjusting a temperature in experimental room. The main advantage of such method is no needs for manual setup, the control system is naturally self-learned from the user.

Klíčová slova anglicky

Pattern recognition, Naive Bayessian Filter, Machine learning, Smart thermostat.

Vydáno

08.06.2016

Nakladatel

Brno University of Technology

Místo

Brno

ISBN

978-80-214-5365-4

Kniha

Mender 2016 - 22nd International Conference on Soft Computing

Strany od–do

149–152

Počet stran

4

BIBTEX


@inproceedings{BUT127362,
  author="Stanislav {Věchet} and Jan {Hrbáček} and Michal {Růžička} and Petr {Mašek} and Jiří {Krejsa},
  title="Behavioral Patterns Identification in Smart-Home Application",
  booktitle="Mender 2016 - 22nd International Conference on Soft Computing",
  year="2016",
  month="June",
  pages="149--152",
  publisher="Brno University of Technology",
  address="Brno",
  isbn="978-80-214-5365-4"
}