Title :
Probabilistic approach of a decision tree to fault diagnosis on power distribution lines
Author :
Togami, M. ; Abe, N. ; Kitahashi, T. ; Ogawa, H.
Author_Institution :
Nagoya MFG, Japan
Abstract :
A method for creating decision trees, which the authors propose in this paper, can distinguish power distribution line fault examples whose attribute value is disjunction in area even when value range of two optional categories are not disjoint to each other. This method can estimate which category a given example belongs to and how much the possibility is.
Keywords :
decision trees; fault diagnosis; power distribution faults; power distribution lines; power distribution reliability; decision tree; fault diagnosis; power distribution lines; probabilistic approach; Artificial neural networks; Classification tree analysis; Conductors; Decision trees; Diagnostic expert systems; Distributed computing; Fault diagnosis; Machine learning; Power distribution lines; Power system faults;
Conference_Titel :
Electric Power Engineering, 1999. PowerTech Budapest 99. International Conference on
Conference_Location :
Budapest, Hungary
Print_ISBN :
0-7803-5836-8
DOI :
10.1109/PTC.1999.826560
