基于云模型的入侵检测日志数据特征选择算法

Feature selection of the intrusion detection logs based on particle swarm optimization

提出一种基于云模型和粒子群优化算法相结合的数据特征选择算法,实现了特征属性的有效约简,降低了特征属性的冗余性,有助于提高日志数据的评估速度.实验结果表明,提出的FSA-CP算法在选择准确率和算法收敛性等方面具有一定优势.

We propose an approach called FSA-CP based on cloud model and particle swarm optimization（PSO） to realize feature selection by achieving an effective reduction of feature attributes,which solves the problem of feature redundancy and helps improve the speed of the evaluation.Experimental results show that the FSA-CP algorithm has an advantage in accuracy and convergence.