摘要
生物是解决问题的能手,通过模拟生物进化过程可以解决很多非线性问题.基于误用的入侵检测最大的弱点就是无法检测未知的攻击.为了使入侵检测系统增加自学习能力,从海量的审计日志中搜索出最优的攻击脚本子集,引入了基因算法进行启发式搜索.并通过仿真实验,达到动态优化脚本库的效果.
Through simulating the evolutionary process,many non-linear problems can be resolved.The biggest shortcoming of misuse detection is the lack for the ability to detect the unknown intrusions.Genetic algorithm was put into use ,in order to enhance the machine learing so as to find out the most hazardous attack subsets among huge data recorded in audit trail.The data was analyzed via the emulation test,and in the end the scenario set was optimized dynamically.
出处
《应用科技》
CAS
2004年第2期41-43,共3页
Applied Science and Technology
