Supervised Learning Approach to Detect and Evaluate Malicious Inside the Cloud Environment

Abstract

SaaS, PaaS, and IaaS platforms are available to meet any organization’s IT needs. As a result, all industries are moving their infrastructure to the cloud. To combat this, attackers frequently attack the environment by exploiting a vulnerability in the infrastructure due to its distributed nature and dynamic configuration. Various security frameworks are widely used on customer and cloud provider premises. The attacks, however, are becoming more frequent. To improve the security on traditional frameworks, this paper introduced a security framework for evaluating the network packet behavior. It provides a solution for a variety of rapidly growing network attacks, and it will help to detect harmful network activities. Several classification algorithms of Machine Learning are used in this paper, including Random Forest (RF), Decision Tree (DT), Naive Bayes (NB), Support Vector Machine (SVM), and K-nearest neighbor (KNN). We tested our method on the UNSW-NB15 standard dataset. The results show that our method provides more accuracy and better result.