Review of Detection Denial of Service Attacks using Machine Learning through Ensemble Learning

Abstract

Today's network hacking is more resource-intensive because the goal is to prohibit the user from using the network's resources when the target is either offensive or for financial gain, especially in businesses and organizations. That relies on the Internet like Amazon Due to this, several techniques, such as artificial intelligence algorithms like machine learning (ML) and deep learning (DL), have been developed to identify intrusion and network infiltration and discriminate between legitimate and unauthorized users. Application of machine learning and ensemble learning algorithms to various datasets, consideration of homogeneous ensembles using a single algorithm type or heterogeneous ensembles using several algorithm types, and evaluation of the discovery outcomes in terms of accuracy or discovery error for detecting attacks. The survey literature provides an overview of the many approaches and approaches of one or more machine-learning algorithms used in various datasets to identify denial of service attacks. It has also been shown that employing the hybrid approach is the most common and produces better attack detection outcomes than using the sole approaches. Numerous machine learning techniques, including support vector machines (SVM), K-Nearest Neighbors (KNN), and ensemble learning like random forest (RF), bagging, and boosting, are illustrated in this work (DT). That is employed in several articles to identify different denial of service (DoS) assaults, including the trojan horse, teardrop, land, smurf, flooding, and worm. That attacks network traffic and resources to deny users access to the resources or to steal confidential information from the company without damaging the system and employs several algorithms to obtain high attack detection accuracy and low false alarm rates.