Reliability assessment of complex networks using rules extracted from trained ANN and SVM models

This paper describes the application of hybrid intelligent system (HIS) to extract rules from two machine learning approaches: neural networks (NN) and support vector machine (SVM). The experimentation is based on the TREPAN algorithm. TREPAN, a well-known technique developed originally to extract linguistic rules from a trained artificial neural network, is modified to cope with SVM models. An example related to the reliability assessment of a 21-links network and its excellent performance results is presented.