An Investigation Through Stochastic Procedures for Solving the Fractional Order Computer Virus Propagation Mathematical Model with Kill Signals

In this study, the numerical investigations through the stochastic procedures for solving a class of fractional order (FO) computer virus propagation (CVP) mathematical model with kill signals (KS), i.e., CVP-KS is presented. The KS gets alert about those viruses, which can be infected through the computer system to decrease the virus propagation danger. The mathematical model of the CVP-KS is based on the SEIR-KS model. The focus of these investigations is to present the numerical solutions of the FO-SEIR-KS model using the sense of Levenberg-Marquardt backpropagation scheme (LMBS) together with the neural networks (NNs), i.e., LMBS-NNs. The use of the one dynamic of the other makes the model nonlinear. Three different FO values have been used to check the performances of the designed scheme for this FO-SEIR-KS nonlinear mathematical model. The statics used in this study is 80%, 10% and 10% for training, testing and certification for solving the FO-SEIR-KS nonlinear mathematical model. The numerical simulations are performed through the stochastic LMBS-NNs scheme for solving the FO-SEIR-KS nonlinear mathematical model. The obtained results will be compared with the design of database reference solutions based on the Adams-Bashforth-Moulton. In order to accomplish the validity, capability, consistency, competence and accuracy of the LMBS-NNs, the numerical results using the error histograms, regression, mean square error, state transitions and correlation have been provided.

Automated summary

This study presents numerical investigations of a fractional order computer virus propagation mathematical model with kill signals using stochastic procedures. The numerical solutions of the model are obtained using the Levenberg-Marquardt backpropagation scheme and neural networks. The results demonstrate the performance of the scheme for different fractional order values.