基于自适应蚁群聚类的入侵检测

针对蚁群聚类算法在聚类结果中出现部分数据划分不够准确的问题，提出一种基于信息熵调整的自适应混沌蚁群聚类改进算法。该算法通过优化过程中种群的信息熵来衡量演化的程度，自适应地调整信息素更新策略。每一次迭代结束时，使用混沌搜索算子在当前全局最优解附近搜索更好的解。而随着算法的进行，混沌算子搜索范围逐渐缩小，这样混沌算子在蚁群搜索的初期起到防止陷入局部最优的作用，在蚁群搜索后期起到提高搜索精度的作用，从而得到更好的聚类结果。使用KDD Cup 1999入侵检测数据集所作的仿真实验结果表明，聚类效果改进明显，并能有效提高入侵检测的检测率、降低误检率。

Adaptive ant colony clustering method for intrusion detection

For the problem that partial data partition is not accurate enough in clustering results of ant colony clustering algorithm, an improved adaptive chaotic ant colony clustering algorithm, based on information entropy is proposed.The algorithm measures the evolutive degree by optimizing the population information entropy,and adjusts the pheromone update strat- egy adaptively.It uses the chaotic search operator to search better solution near current global optimal solution at the end of each iteration.With progress of the algorithm ~ search range of the chaotic operator is gradually reduced so that chaotic operator avoids falling into local optimum in the initial period and improves search precision in the later period of ant colony search.This leads to better clustering results.Using the KDD Cup 1999 intrusion detection data,simulation results show that the clustering effect improves significantly, and can effectively improve the detection rate of intrusion detection and reduce the false detection rate.