基于邻域粗糙集的网络入侵分类诊断组合模型研究

发布时间：2018-05-20 07:17

本文选题：网络入侵 + 领域粗糙集　；参考：《数学的实践与认识》2017年10期

【摘要】：网络入侵诊断直接影响网络正常运行和安全.针对入侵类型复杂,现有分类诊断模型精度有限的问题,提出一种基于邻域粗糙集的网络入侵分类诊断优化模型.首先,运用邻域粗糙集对网络入侵数据进行条件属性的约简,确定关键属性,然后将其作为训练输入构建相关向量机分类诊断模型,并同时运用遗传算法进行超参数优化,提高模型诊断精度和速度.通过KDDCup99数据集对优化模型性能进行检验,结果表明,组合预测方法精确度高于支持向量机、相关向量机和BP神经网络.组合模型诊断精度高、速度快,具有优异的综合性能.
[Abstract]:Network intrusion diagnosis directly affects the normal operation and security of the network. Aiming at the problem of complex intrusion types and limited accuracy of existing classification and diagnosis models, a network intrusion classification and diagnosis optimization model based on neighborhood rough set is proposed. Firstly, using neighborhood rough set to reduce the conditional attributes of network intrusion data, the key attributes are determined, and then the classification and diagnosis model of correlation vector machine is constructed as the training input, and the genetic algorithm is used to optimize the super-parameters. Improve the accuracy and speed of model diagnosis. The performance of the optimized model is tested by KDDCup99 data set. The results show that the accuracy of the combined prediction method is higher than that of support vector machine, correlation vector machine and BP neural network. The combined model has high diagnostic accuracy, high speed and excellent comprehensive performance.
【作者单位】：信阳职业技术学院数学与计算机科学学院;昆明理工大学质量发展研究院;
【基金】：国家自然科学基金(51066002/E060701) NSFC-云南联合基金资助项目(U0937604)
【分类号】：TP18;TP393.08

【相似文献】