DEA-based data mining for energy consumption

This paper discusses the potential of DEA (data envelopment analysis) for performance evaluation on energy consumption. The purpose is to develop evaluation methodology for finding features of energy provision and consumption from observed data. While there are four evaluation measures (economy, environment, comfort, and risk) in general, this paper focuses on economical and environmental aspects, and selects energy consumption, CO2 discharge and cost as inputs for the first step model. In order to consider the evaluation methodology based on four plants data, this paper has focused on hot-water supply as output. Then the numerical study illustrates the findings from data on energy supply patterns, seasonal effects, and so on.