ENHANCING CYBER THREAT DETECTION: A DEEP NEURAL NETWORK APPROACH ON THE CTU-13 DATASET

This paper presents a novel deep neural network (DNN) model specifically designed to enhance intrusion detection capabilities within the field of cybersecurity. Utilizing the comprehensive and intricate CTU-13 dataset, our model integrates an innovative dual-path architecture designed to optimize the processing and extraction of features critical for the identification of complex cyber-attack patterns. The effectiveness of this model is demonstrated by its remarkable accuracy rate of 97.52% in identifying malicious activities within the dataset, which is notable for its simulation of real-world network complexities. This research represents a pivotal advancement in the realm of cybersecurity, illustrating the profound impact that sophisticated deep learning architectures can have on the enhancement of threat detection and the fortification of information systems against diverse network intrusions. By improving the accuracy and speed of threat detection, our model contributes significantly to the ongoing efforts to secure cyberspace and protect sensitive data from the increasingly sophisticated methods employed by cyber adversaries.