Thomas W. Rouber,  Idalino Drago,  Flávio Miguel Varejão,  Rodrigo Mendonça Queiroga. 

This paper describes a software tool for automatic detection of frauds in power distribution systems. Even with a high portion of noisy and contradictory data we extract discriminative information by appropriate feature extraction and selection. We principally focus on the time series of the power consumption and performance criteria based on the confusion matrix in a classification context.

http://www.lbd.dcc.ufmg.br/colecoes/enia/2005/069.pdf

Processando consulta.