Evaluating the Efficacy of Deep Learning Models in Identifying and Mitigating Cybersecurity Threats and Attacks

Due to rapid advancement in digital infrastructure and Internet cyber security threats increased in sophisticated way that necessitate advanced strategies for identification and mitigation. The present study uses deep learning models that act as powerful tool for identifying and mitigating cyber security threats and attacks that leverages the capability for detection of complex patterns using large datasets. The Efficacy of the deep learning model can be evaluated using the study for identification of different types of cyber security threats such as DoS, DDoS, ransomware, password hacking and information threat. The proposed model identifies and mitigates the threat in less time and it was tested with two different data sets BoT and ToN. The proposed model is measured and evaluated using various performance metrics using cyber security data sets. The study explores the model interpretability and robustness for understanding the potentiality and implications based on cyber security solutions. Study findings shows that the proposed model shows highest accuracy of 99.9% and it is comparatively higher than other existing models in identification and mitigation of threats and attacks.