An Effective Method for Detecting Cyber Attacks on Computer Networks from the NSL-KDD Data Set

¹ Dept. of AIML, Samskruthi College of Engineering and technology(A), Hyderabad, India
² Dept. of S & H, Samskruthi College of Engineering and technology(A), Hyderabad, India
³ Dept. of AIML, Chalapathi College of Engineering and technology(A), Guntur, A.P., India
⁴ Dept. of ME, Sreenidhi Institute of Science and technology(A), Hyderabad, India
⁵ Dept. of IT, Sreenidhi Institute of Science and technology(A), Hyderabad, India

Abstract

Cybercrime is rapidly increasing and exploits various vulnerabilities in these computing environments. Ethical hackers pay more attention to determining vulnerabilities and recommending mitigation methods. Due to the effectiveness of machine learning in solving cybersecurity problems, machine learning is of great importance to cybersecurity. Machine learning models are used to advance the techniques to detect and solve cybersecurity problems. Machine learning methods help detect more cyber attacks more efficiently than other software-oriented techniques, reducing the burden on security analysts. Adaptive methods such as machine learning can improve detection rates. Logistic regression is used to resolve the issue of intrusion identification and a novel research model for intrusion identification. Logistic regression models can fully favor network traffic structure information to capture features more comprehensively. Experimental outcomes show that the algorithm behaves better than traditional methods.