An Adaptive Fault Classification Algorithm Based on Bayesian Network
Lin Chen, Ming Sheng He · International Conference on Electric Information and Control Engineering · 2012
There are many factors affect the behavior of network, those factors are usually dependent on each other with complex association relationship. Non-linear mappings may exist between symptoms and causes of network fault, and the same network faults often have different symptoms at different time, while one symptom is the result of several network faults. There is much correlative information in the network, how to obtain the network fault set most likely causing abnormal network is a challenging topic. The paper firstly present feature selection and learning strategies for network fault classification, and then we proposed an adaptive fault classification algorithm based on Bayesian Network (AFC-Bayes). Compared with previous techniques, experiments have shown that our approach can gain higher classification accuracy in large scale networks.