Análise de desempenho de um modelo de geração de alvos de fiscalização de perdas não técnicas

Abstract

The electrical system is made up of the generation, transmission and distribution sectors, these sectors having losses that can be divided into technical losses and non-technical losses. The objective of the present study is to analyze the performance of a prediction model and identification of outliers referring to non-technical losses. The methodology is based on the use of SPSS Software that, through the data mining process, performs a search for patterns, enabling the appointment of consumers in a situation of probable irregularity. Through the analysis of the assertiveness of the current model, it is sought to improve it and create a new predictive model in order to achieve more accurate results.