CAJ | 학술논문

为了提高网络入侵检测的正确率，提出一种精英选择策略粒子群算法（EPSO）优化相关向量机（RVM）的网络入侵检测模型（EPSO-RVM）。将相关向量机的参数编码成粒子，将入侵检测正确率作为粒子群搜索的目标，通过粒子群算法对参数优化问题进行求解，并引入精英选择策略增强粒子群算法的全局搜索能力，根据最优参数建立基于RVM的入侵检测模型，采用KDD99数据集对其性能测试，结果表明，相对于对比模型，EPSO-RVM较好地解决了相关向量机参数优化难题，提高了网络入侵检测的正确率。
위료제고망락입침검측적정학솔，제출일충정영선택책략입자군산법（EPSO）우화상관향량궤（RVM）적망락입침검측모형（EPSO-RVM）。장상관향량궤적삼수편마성입자，장입침검측정학솔작위입자군수색적목표，통과입자군산법대삼수우화문제진행구해，병인입정영선택책략증강입자군산법적전국수색능력，근거최우삼수건립기우RVM적입침검측모형，채용KDD99수거집대기성능측시，결과표명，상대우대비모형，EPSO-RVM교호지해결료상관향량궤삼수우화난제，제고료망락입침검측적정학솔。
In order to improve the detection precision of network intrusion, this paper proposes a novel network intrusion detection model based on relevance vector machine optimized by particle swarm optimization algorithm with elite election strategy. The parameters of relevance vector machine are encoded into particles, and network intrusion detection rate is taken as the search goal of particle swarms, and then particle swarm optimization algorithm is used to solve the parameters optimi-zation problem in which elite election strategy is introduced to improve the search performance. Network intrusion detec-tion model is established based on the optimal parameters of relevance vector machine, and the simulation experiments are carried out on the KDD99 dataset. The simulation results show that the proposed model has solved parameters optimi-zation problem of relevance vector machine and improved intrusion detection rate compared with reference models.