CYBER THREAT DETECTION BASED ON ARTIFICIAL NEURAL NETWORKS

CYBER THREAT DETECTION BASED ON ARTIFICIAL NEURAL NETWORKS

1P. Shiva Reddy, 2K. Sai Nitej, 3Amancharla Lakshmi Kausika, 4 Dr. M. Rama Chandra

1 B.Tech Scholar, Dept. Computer Science & Engineering,SNIST,Hyderabad-501301,India

2 B.Tech Scholar, Dept. Computer Science & Engineering, SNIST,Hyderabad-501301,India

3 B.Tech Scholar, Dept. Computer Science & Engineering,SNIST,Hyderabad-501301,India

4Associate Professor, Dept. Computer Science & Engineering, SNIST, Hyderabad-501301, India

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - To address the difficulty of an effective automated detection technique in cyber security, this research study provides an AI technique for cyber-threat detection utilising artificial neural networks. For enhanced cyber-threat detection, the suggested strategy involves transforming collected security events to individual profiles and employing traditional machine learning-based algorithms such as ANN, k-NN, RF, NB, and DT. To aid security analysts in responding quickly to attacks, the system seeks to discern between true and false positive signals. To measure performance, the authors ran trials with four datasets and five traditional machine-learning approaches. The results show that artificial neural networks (ANNs) have shown great promise in cyber threat detection using event logs, achieving 98.0% accuracy in identifying various types of attacks and can be used as a learning-based model for network intrusion detection in real-world scenarios.

Keywords: Cybersecurity, Cyber-threat detection, Artificial intelligence, Artificial neural networks, traditional machine-learning methods