作者: Yuesheng Gu , Yanpei Liu , Hongyu Feng
DOI: 10.1007/978-3-642-27452-7_59
关键词:
摘要: It’s very important to detect the network attacks protect information security. The intrusion patter identification is a hot topic and using artificial neural networks (ANN) provide intelligent recognition has been received lot of attentions. However, detection rate often affected by structure parameters ANN. Improper ANN model design may result in low precision. To overcome these problems, new approach based on improved genetic algorithm (GA) BPNN classifiers proposed this paper. GA used energy entropy select individuals optimize training procedure BPNN, satisfactory with proper parameters. efficiency method was evaluated practical data. experiment results show that offers good rate, performs better than standard GA-BPNN method.