A comparative study of classical and neural network based optimal control in dengue transmission dynamics.

This study develops and analyses a mathematical model that represents dengue transmission between human and mosquito populations. The human population is divided into susceptible, infected, and recovered classes, while the mosquito population is divided into susceptible and infected classes. This model captures interaction between humans and vectors with a number of significant epidemiological parameters governing the spread of disease. In devising effective intervention strategies, two optimal control techniques are considered. The first technique applies Pontryagin's Minimum Principle for obtaining analytical optimal control conditions, whereas the second one makes use of an Artificial Neural Network structure that approximates the dynamics of the system and optimizes the control process by a learning-based strategy. Numerical simulations indicate that both methods reduce significantly the infected populations and vectors, and the ANN-based method is more efficient and flexible when dealing with the nonlinear dynamics of the system. The results highlight the importance of integrating traditional mathematical analysis with state-of-the-art computational intelligence methods to facilitate dengue prevention and control.