An Enhanced Network Security using Machine Learning and Behavioral Analysis with Voting Classifier

Abstract

With exponential growth in Internet use, the prevalence of cyber attacks has increased sharply, required by robust
intrusion detection systems to secure network. This study is a new approach of supervised machine learning aimed
at increasing the network security by accurately classification of network traffic as harmful or benign. The model
that uses a mixture of algorithms of supervised learning and function selection techniques maximizes the detection
rate by identifying the appropriate functions and using advanced algorithms. The model's performance evaluation
uses the NSL-KDD data file, recognized benchmark for algorithms of network traffic classification.