Machine Learning Based Intrusion Detection System in Cloud Environment

Abstract

Attacks and intrusions on computer networks often have different characteristics and behaviors that require professional help. The number of attacks is growing in line with the development of computer networks. In fact, expert knowledge is declining over time and should be reviewed and made available in the system, making it necessary to hire an experienced person. Cloud computing is an advancement in IT technologies that provide users with the most up-to-date, highly sought-after virtual services with high flexibility, low infrastructure costs, and minimal maintenance. Protection against network intruders is one of the most important security challenges for cloud computing, as it affects the privacy, availability, and integrity of cloud services. Because cloud computing is a shared environment, it can be vulnerable to various threats. Because building strong access systems are essential to maintaining cloud security, but it remains an obscure goal and a major challenge due to the growing number of cyber-attacks. In this study, a machine-learning based approach was used to propose an intrusion detection system (IDS) based on a network based cloud computing model. During the previous processing phase, we used a feature selection algorithm from the CICDDoS2019 database to select features. We have used the Naive Bayes Classifier, Decision Tree Classifier, Supporting Classifier, Logistic Regression, and Random Forest Classifier, all of which are well-known categories. Simulation results when compared to these five different classifiers, random forest algorithm show improvements in overall accuracy, precision, recall and F1 score. We explain how various characteristics of machine learning techniques can be used to build efficient IDS. The goal of this research is to develop a new method based on intrusion detection systems and their varied designs for improving intrusion detection accuracy in cloud computing.