Download PDFOpen PDF in browserMalware Detection Using Machine LearningEasyChair Preprint 1273031 pages•Date: March 27, 2024AbstractIn today’s digital landscape, safeguarding computer systems against malicious software, or malware, is of paramount importance. This project delves into the realm of malware detection using advanced Machine Learning techniques. The dataset comprises a diverse set of network traffic attributes, providing a rich foundation for analysis and model training. The dataset encompasses a total of 12,989 entries with 43 distinct features. These features encapsulate crucial information about network activities, including duration, protocol type, services utilized, flags, and numerous others. Each entry is associated with a unique identifier (Id) and is labeled with a type of attack, enabling supervised learning for classification. The prediction classes, representing different types of attacks, are defined as follows: ”ipsweep,” ”satan,” ”portsweep,” ”back,” and ”normal.” Each of these classes represents a distinct category of network behavior, ranging from suspicious and potentially harmful activities to benign, legitimate traffic. In conclusion, this project showcases a comprehensive exploration of malware detection through Machine Learning, utilizing a diverse dataset of network measures and safeguard digital assets against malicious attacks. Keyphrases: machine learning, network traffic analysis, supervised learning
|