Title :
New Insights on Nontechnical Losses Characterization Through Evolutionary-Based Feature Selection
Author :
Ramos, Caio César Oba ; De Souza, André Nunes ; Falcão, Alexandre Xavier ; Papa, João Paulo
Author_Institution :
Dept. of Electr. Eng., Univ. of Sao Paulo, Sao Paulo, Brazil
Abstract :
Although nontechnical losses automatic identification has been massively studied, the problem of selecting the most representative features in order to boost the identification accuracy and to characterize possible illegal consumers has not attracted much attention in this context. In this paper, we focus on this problem by reviewing three evolutionary-based techniques for feature selection, and we also introduce one of them in this context. The results demonstrated that selecting the most representative features can improve a lot of the classification accuracy of possible frauds in datasets composed by industrial and commercial profiles.
Keywords :
evolutionary computation; learning (artificial intelligence); particle swarm optimisation; pattern classification; power distribution economics; power engineering computing; security of data; commercial profile; dataset fraud classification; evolutionary-based feature selection; illegal consumers; industrial profile; machine learning; nontechnical losses characterization; power distribution systems; Accuracy; Context; Force; Optimization; Search problems; Training; Vectors; Feature selection; gravitational search algorithm; harmony search; nontechnical losses; optimum-path forest; particle swarm optimization; pattern recognition;
Journal_Title :
Power Delivery, IEEE Transactions on
DOI :
10.1109/TPWRD.2011.2170182