Targeting of the captive consumer market for high and medium voltage of an electric power distributor company: application of the Kohonen Self-Organization Map for discovery of industry delinquency patterns

Abstract

The incentivized regulation regime, in which electricity distribution agents operate, is frequently threatened by delinquency that affects the financial-economic balance of the distribution market. In light of this, it is essential that distribution agents have a methodological tool that allows identifying preventively customer payment profiles and thus formulating strategies for tackling delinquency in this market. The distribution market is characterized as a natural monopoly and by the predominance of captive consumers. This paper proposes the use of an unsupervised neural network for grouping consumer units according to information stored in a database and then obtaining a pattern of delinquency by exploratory data analysis, considering the similarities of the created groups. The results show that this study is effective and suitable for discovering patterns in others markets.