Innovative time series forecasting: auto regressive moving average vs deep networks
Growing interest in meaningful indicators extraction from the huge amounts of data generated by energy efficient buildings instrumentations has led to focusing on so called smart analysis algorithms. This work proposes to focus on statistical and machine learning approaches that make use only of available data to learn relationships, correlations and dependencies between signals. In particular, time series forecasting is a key indication to anticipate, prevent and detect anomalies or unexpected behaviors. We propose to compare performances of a classical Auto Regressive Moving Average (ARMA) approach to a Deep Highway Network on time serie forecasting only making use of past values of the serie. In recent years, Deep Learning has been extensively used for many classification or detection tasks. The complexity of such models is often an argument to discard such approaches for time serie prediction with regard to more common approaches performances. Here we give a first attempt to evaluate benefits of one of the most up to date Deep Learning model in the literature for time serie prediction.
sustainability, buildings, time series forecasting, Auto Regressive Moving Average (ARMA), deep networks
C45 , C53
FundingThis research was supported by the PERFORMER project. Project funding from the European Union’s Seventh Programme for research, technological development and demonstration under grant agreement No 609154.
This is an open access issue and all published articles are licensed under a
Creative Commons Attribution 4.0 International License