Characterization of Non-Technical Loss in Consumers of the Electric Distribution System

This paper presents the results of an exploratory analysis performed on a set of consumer characteristics of the electric distribution system. The main objective was to identify the most important attributes for recognizing consumers associated with non-technical losses. To achieve this, a hybrid approach was employed, combining heuristic analysis with statistical techniques, including correlation coefficients, mutual information, and scatter plots. As a result of the study, it was possible to reduce the initial set of 385 attributes to just 20, which were also ranked in order of importance. The selected attributes contain relevant information for classifying consumers and will later be used as input in a predictive model. The findings of this paper can help energy companies in understanding which consumer information should be monitored to detect non-technical losses.