A Data-Driven Hybrid Artificial Neural Network and Ordinary Differential Equation Model for Nonlinear Dengue Transmission Dynamics

Female Aedes aegypti mosquitoes are the primary vectors responsible for transmitting the dengue virus. Due to climate change, rapid urbanisation, population growth, and seasonal rainfall, dengue continues to spread globally and has been identified by the World Health Organisation as one of the most significant public health threats. This study presents a hybrid modeling framework for dengue transmission by integrating artificial neural network with compartmental ordinary differential equation model to capture its complex and nonlinear dynamics more precisely. Fundamental properties of the model, including positivity, boundedness, and equilibrium points, are analytically examined. The basic reproduction number is derived using the next-generation matrix approach, which determines whether the disease will die out or persist in population. Model parameters are estimated using a least-squares method based on reported dengue data from Liberia covering the period 1990–2015. Furthermore, an artificial neural network is integrated with the Ordinary Differential Equation model to effectively learn complex nonlinear relationships from reference data, improving model accuracy, robustness, and predictive capability compared to traditional estimation methods. Sensitivity analysis is performed to identify parameters that significantly influence the transmission dynamics of dengue. Numerical simulations demonstrate that the combined Artificial Neural Network–Ordinary Differential Equation framework provides reliable insights into dengue dynamics and supports effective disease control strategies.