实数向量型阴性选择算法的改进

An Improved Algorithm of Real-valued Vector of Negative Selection

通过对实数向量型阴性选择算法的分析,提出了检测器空间覆盖率的概念,用它作为估计检测器数量的一项理论依据,将这个估计值引入到实数向量型阴性选择算法中,控制检测器的生成,同时对检测器采取了新的变异操作。实验表明,这一改进在保证算法检出率的同时,又可降低误报率。

Based on the study of Algorithm of Real-valued Vector of Negative Selection, the concept of space coverage of detectors is introduced and used as a theoretical basis for the estimated number of detector. We apply this estimated number to the algorithm to control the produce of detectors. Then we adopt new mutation trend to detectors in the Algorithm. From the results of testing on the improved step, the new algorithm can ensure preferable detection rate and low false alarm rate.