Journal
COMPUTER SCIENCE-RESEARCH AND DEVELOPMENT
Volume 31, Issue 3, Pages 141-148Publisher
SPRINGER HEIDELBERG
DOI: 10.1007/s00450-014-0294-4
Keywords
Smart meter data (SMD); Household classification; Supervised machine learning; Feature selection (FS); Feature filtering; Extended CLASS (eCLASS)
Categories
Ask authors/readers for more resources
Knowing household properties, such as number of persons per apartment, age of housing, type of water heating, etc. enables energy consultants and utilities to develop targeted energy conservation services. Load profiles captured by smart power meters, can-besides several other applications-be used to reveal energy efficiency relevant household characteristics. The goal of this work is to develop methods of supervised machine learning that deduce properties of private dwellings using consumption time series recorded in 30-min intervals. The contribution of this paper to the state of the art is threefold: we quadruplicate the number of features that describe power consumption curves to preserve classification relevant structures, indicate dimensionality reduction techniques to reduce the large-scale input data to a set of few significant features and finally, we redefine classes for some properties. As a result, the classification accuracy is elevated up to 82 %, while the runtime complexity is significantly reduced. The classification quality that can be achieved by our eCLASS methodology renders personalized efficiency measures in large-scale practical settings possible.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available