Analysis of Network Traffic Using MRF Algorithm in Machine Learning

R.V. Ritthish

S. Logamoorthy

T. Batsha

K. Balamurugan

Keywords: Machine Learning, Deep Learning.


Digital risks are developing quickly thanks to the development of the Internet, and the outlook for digital security isn't always promising. Artificial intelligence (AI) and deep learning (DL) techniques are used to assess local area interruption discovery and provide a brief but informative description of each ML/DL strategy. Based only on their connections to other people or places, the papers discussing each technique have been gathered, reviewed, and summarized. Since data are so important to ML/DL methods, they illustrate some of the most often used local datasets, discuss the use cases for ML/DL in digital assurance, and propose principles for heading-focused headings. The KDD data set is frequently recognized as a standard in studies of intrusion detection methods. Studies into the data used for tutoring and examination of the discovery adaption are both of utmost relevance since improved data quality can boost unconnected interruption detection. There is a lot of work being done to improve interruption location methods. This task provides an analysis of the KDD data set with regard to four examples: Basic, Content, Traffic, and Host. All data attributes may be categorized using MODIFIED RANDOM FOREST (MRF). For an intrusion detection system (IDS), the evaluation is conducted using the Detection Rate (DR) and False Alarm Rate (FAR) metrics. As a consequence of this experimental analysis of the data set, the dedication of all examples of attributes on DR and FAR is demonstrated, which may assist improve the reasonableness of the data set to obtain the highest DR with negligible FAR.