REsearch and Analysis of Network Behavior Based on Traffic

With the continuous development of the network and the continuous development of various applications, it brings convenience to human beings. However, it has also led to an increase in diverse security issues. For instance, complex and varied network attacks continuously pose threats to people's online activities. In order to detect and prevent such network attacks, mitigate their negative impact, and safeguard individuals' privacy and property, this paper focuses on analyzing network traffic. This study employs the K-Nearest Neighbors (KNN) classification algorithm to identify six types of network behaviors, including normal network behavior, monitoring, denial of service, and unauthorized access. Initially, the KDD99 dataset is preprocessed, followed by an analysis of its characteristics. Subsequently, machine learning techniques are utilized to detect network behaviors within the network traffic. Finally, the network traffic is evaluated to determine various network behaviors, thereby achieving the objective of behavior identification within network traffic. The effectiveness of the proposed method in this paper is verified using the publicly available KDD99 dataset. The experimental results demonstrate that the approach presented in this paper can accurately detect different network behaviors, achieving an impressive accuracy rate of 97.0
^%
with a low false positive rate of only 3.0%.