A Mathematical Real Analysis on 2D Connection Spaces for Network Cyber Threats: A SEIAR-Neural Network Approach

This article examines this problem by deploying the mathematical modelling, specifically compartmental SEIAR model to simulate and analyse cyberattack on a number of devices linking with a server. Using epidemiological modelling techniques to understand cyber threat, this approach, also known as malware contagion research spans from the early 2000's, where researchers applied principles of epidemic theory and benchmark methods used by National Institute for Computer Science. This research explores a method of computational work on an Artificial Neural Networks with the system of ODE that describes SEIAR model in Distributed Denial of Service attacks. The study uses a feed-forward neural network trained with back-propagated Levenberg-Marquardt algorithmascular bed dysfunction. We choose a machine-learning-based approach here because it is known to work well with linear scenarios and has faster convergence rate; hence we use this as an optimisation method. The basis FNN-BLMA results are then evaluated with the validated solutions obtained using RK-4optimizer in order to find out about how much accurate and effective is said method. The agreement is very good, with only small absolute errors between the approximate and reference solutions. We demonstrate the strength of our design strategy with by showing convergence analysis, error histograms and regression on each differential equations in the SEIAR model. This work emphasizes the potential of machine learning methodology, ANNs particularly, in building robust defence countermeasures to epidemic cyber security threats. The results reveal that the FNN-BLMA methodology could achieve high accuracy by predicting the properties of real-world cyberattacks under different conditions, leading to a better performance on mitigating strategies.