Unmasking unbalanced network traffic by intruders based on Machine-learning Algorithms.

Abstract

Obtrusion unmasking involves identifying and detecting malicious or anomalous network activity. Imbalanced network traffic refers to a scenario where the ratio of normal to abnormal traffic is significantly skewed, with one type of traffic significantly outweighing the other. Machine learning and deep learningtechniques can be used to analyze network traffic and identify patterns and characteristics that may indicate the presence of obtrusions or anomalies.  These techniques can be applied to various types of network data, including packets, flows, and logs, and can be trained on large datasets to improve their accuracy and effectiveness. For the purpose of predicting the intrusion in imbalance network traffic, this research employs a variety of categorization techniques. The classification algorithms are Random Forest, Support Vector Machine, convolution neural network, and principal component analysis. The dataset, which included several meteorological parameters, was obtained via the UCI repo. With a testing data ratio of 70:30. Based on precision, accuracy, recall, f1-score, and execution time, the efficacy of the classification algorithms was assessed.