We consider the problem of automated anomaly detection for building level heat load time series. An anomaly detection model must be applicable to a diverse group of buildings and provide robust results on heat load time series with low signal-to-noise ratios, several seasonalities, and significant exogenous effects. We propose to employ a probabilistic forecast combination approach based on an ensemble of deterministic forecasts in an anomaly detection scheme that classifies observed values based on their probability under a predictive distribution. We show empirically that forecast based anomaly detection provides improved accuracy when employing a forecast combination approach.
@article{arxiv.2107.10828,
title = {Probabilistic Forecast Combination for Anomaly Detection in Building Heat Load Time Series},
author = {Mario Beykirch and Tim Janke and Imed Tayeche and Florian Steinke},
journal= {arXiv preprint arXiv:2107.10828},
year = {2021}
}
Comments
Accepted in the proceedings of ISGT-Europe 2021 to be published by IEEE