Dynamical analysis of a fractional Maize Streak Disease model with enhanced predictive capabilities

Maize Streak Disease (MSD), caused by the Maize Streak Virus (MSV) and transmitted by leafhoppers of the genus Cicadulina, poses a significant threat to global maize production. This study presents a novel fractal-fractional Caputo derivative-based model to analyze the transmission behaviors of MSD and evaluate the controllability in terms of insecticide interventions. The model incorporates susceptible, insecticide-treated, exposed, infected, and recovered maize plant compartments. The existence and uniqueness of solutions are established using fixed-point theory, and Ulam-Hyers stability is investigated. Numerical simulations, performed using the fractional Adams-Bashforth method, demonstrate the impact of varying fractional orders Λ and fractal dimensions χ on disease dynamics along with their sensitivity analysis. Our results approximated the controlling through insecticides in reducing disease transmission and provide insights for optimizing control strategies. The fractal-fractional approach offers a more comprehensive understanding of MSD dynamics compared to classical integer-order models. Additionally, a deep neural network method was applied to perform a more detailed analysis of the MSD model using the parameters data of the MSD model. We divided the data set into three categories. The dataset was divided into training, testing, and validation sets. Two activation layers were used: a tanh function in the first layer and a linear function in the second.