Fractional-order Modeling of Infectious Diseases: a Stochastic Neural Network Procedure to Deal with Vaccination and Awareness Strategies

The aim of the work is to solve the fractional order infectious disease model with the awareness and vaccination effects by executing reliable neural network strategies. Fractional kinds of derivatives perform higher efficiency and accuracy in comparison with the derivatives of integer kinds. The fractional order infectious disease model with the awareness and vaccination effects is separated into susceptible class, vaccinated class, infected class, quarantined class, and removed class. A construction of the proposed neural network is accomplished by a single layer construction with log-sigmoid transfer function together with 24 neurons. The model is trained using the Adam optimizer along with the Bayesian regularization, a reliable solver to perform the results of nonlinear systems. The dataset obtained between 0 and 1 with the step size of 0.01, which is divided into three states: validation 10