Title :
Use of reliability measures to improve the performance of fuzzy ARTMAP networks
Author :
Amuhalli, P.R. ; Udpa, L. ; Udpa, S.S.
Author_Institution :
Mater. Assessment Res. Group, Iowa State Univ., Ames, IA, USA
Abstract :
Neural network based signal classification systems are being applied increasingly in nondestructive evaluation to solve the inverse problem. In general, two issues not usually addressed are (i) estimation of reliability measures of the network decision and (ii) ability of the network to learn and improve its performance with time. This paper presents a signal classification system using the fuzzy ARTMAP network. Fuzzy logic based reliability measures are developed for the fuzzy ARTMAP network and used as a feedback for retraining the network to improve its performance. The performance of the algorithm is demonstrated using ultrasonic data obtained from the inspection of welds in nuclear flow plant piping
Keywords :
ART neural nets; fission reactor materials; fission reactor safety; fuzzy neural nets; inspection; inverse problems; learning (artificial intelligence); nuclear engineering computing; nuclear power stations; power system reliability; recurrent neural nets; signal classification; ultrasonic materials testing; US data; feedback; fuzzy ARTMAP networks; fuzzy logic based reliability measures; inverse problem; learning ability; network decision reliability measure estimation; network retraining; nondestructive evaluation; nuclear flow plant piping; reliability measures; signal classification systems; ultrasonic data; weld inspection; Fuzzy logic; Fuzzy systems; Inspection; Inverse problems; Neural networks; Pattern classification; Power generation; Subspace constraints; Ultrasonic variables measurement; Welding;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.830802
