作者: Adel Issa , Adnan Abdulazeez
DOI: 10.33899/CSMJ.2011.163644
关键词: Intrusion detection system 、 Data mining 、 Backpropagation 、 Artificial neural network 、 Engineering 、 Sigmoid activation function
摘要: Vulnerabilities in common security components such as firewalls are inevitable. Intrusion Detection Systems (IDS) used another wall to protect computer systems and identify corresponding vulnerabilities. The purpose of this paper is use Backpropagation algorithm for IDS by applying bipolar input "input represented (1, -1)", sigmoid activation function. KDD Cup 99 dataset paper. Number train 4947 connection records, number test 3117 records. results the proposed method show that PSP 88.32 CPT equal 0.286.