论文部分内容阅读
针对网络异常检测领域存在的漏报率和误报率较高的问题,提出一种基于Kmeans聚类的网络流量异常检测方法。选择了多个不同维度上的特征;计算各维特征在滑动窗口中的局部均值偏差,以保证在实时动态变化的网络中的检测准确度;利用由K-means聚类算法产生的检测模型对各维特征进行综合评判,有效地降低了漏报率和误报率。在网络流量数据集上对所提方法进行了验证并和已有方法进行了对比,所提方法在精度和效率方面取得了较好的实验效果。
Aiming at the problem of high false negative rate and false negative rate in network anomaly detection, a Kmeans clustering based network traffic anomaly detection method is proposed. Selecting features in different dimensions; calculating the local mean deviation of each feature in the sliding window to ensure the detection accuracy in the real-time dynamic network; using the detection model generated by the K-means clustering algorithm Comprehensive evaluation of the features of various dimensions, effectively reducing the false negative rate and false positive rate. The proposed method is verified on the network traffic dataset and compared with the existing methods. The proposed method has achieved good experimental results in accuracy and efficiency.