一种提高检测率的免疫优化模型

An Immune Principle and Optimization Model Increasing Detection Rate

文章建立了一种基于免疫原理的提高检测率的优化模型。通过提出匹配强度概念,给出了检测子之间的重叠和检测集与自我集的重叠的计算方法。进一步用最小化检测子之间的重叠代替最大化检测集的覆盖,用检测集与自我集的重叠代替自身耐受约束,建立了计算检测集分布的优化模型,避免了在检测集的覆盖计算中使用并集公式和蒙特卡罗方法。选择经典的Iris数据集进行实验,结果表明,基于免疫优化模型的二进选择算法产生的检测集的覆盖大于线性算法和贪婪算法产生的检测集的覆盖,提高了入侵检测等异常检测的检测率。

This paper proposes an immune optimization model to increase intrusion detection rate. Via puting forward the concept of match strength, overlapping among detectors and overlapping between detectors and self set are defined, and instead of maximizing the cover of detectors and restricted by self tolerance, minimizing overlapping among detectors and overlapping between detectors and self set is used separately. Optimization model of distribution of detectors is established, which avoids the difficult to use union of set formula and Monte Carlo method directly. Iris data set is chosen to carry through experiments, the results show that the coverage of detector set generated by anomaly detection method based on immune principle and optimization model is larger than that generated by linear time detector generating algorithm and greedy detector generating algorithm, and can increase detection rate when applying in intrusion detection.