Título: A Framework for Evaluating Data Mining Techniques Applied to Power Quality
Autores: Pires, Yomara; Santos, Andreia; Borges, Jose; Carvalho, Antonio; Nunes, Marcus Vinicius Alves; Santoso, Surya; Klautau, Aldebaro
Resumo: Real-world power quality (PQ) events usually are infrequent and composing a dataset of significant size is a difficult task. Using simulators to generate data in wellcontrolled conditions is a very useful and popular option. This work presents a framework based on the ATP (Alternative Transient Program) simulator, which allows for evaluating data mining techniques applied to PQ events, such as fault classification using support vector machines (SVM) and neural networks.
Palavras-chave: Power quality; pattern recognition; Time series classification; fault classification; data mining
Código DOI: 10.21528/CBRN2005-193
Artigo em PDF: CBRN2005_193.pdf
Arquivo BibTex: CBRN2005_193.bib