Aiming at the high-dimensional data generated by intrusion detection systems and SVM parameter optimization problems,the paper puts forward a network intrusion detection model with genetic algorithm synchronous selecting feature and SVMparameters.At first the feature subsets and SVMparameters are coded as chromosome,and the network intrusion detection categorized accuracy is taken as grouped individual fitness degree value.Then depending on the global search ability of the genetic algorithm,it synchronously finds out feature combi-nations that most influence the categorization algorithm and the SVMoptimal parameters.At last it uses KDD99 datasets to carry out simula-tion experiments.Results show that the model can quickly find the optimal feature subset and SVMparameters and the network intrusion de-tection accuracy ratio is improved,so that it is regarded as a good network intrusion detection algorithm.%针对入侵检测系统产生的高维数据和支持向量机参数优化问题，提出一种遗传算法同步选择特征和支持向量机参数的网络入侵检测模型。首先将特征子集和支持向量机参数编码成染色体，将网络入侵检测的分类准确率作为种群个体的适应度值，然后通过遗传算法的全局搜索能力，同步找到对分类算法最有影响的特征组合和支持向量机最优参数，最后采用 KDD99数据集进行仿真实验。结果表明，该模型可以快速找到最优特征子集和支持向量机参数，提高了网络入侵检测正确率，是一种较好的网络入侵检测算法。
展开▼
