On Sensitivity Analysis of A Nonlinear Mathematical Approach of Alcohol Consumption: A Piecewise Fractional and ANN Approach

One of the greatest challenges in public health is alcoholism, which impacts people from different socioeconomic backgrounds alike. The current research presents a mathematical model comprised of five compartments to depict the development of alcohol consumption and addiction. Nonlinear behaviour that is complex is noticed in actual drinking patterns, and memory effects are captured through the model's incorporation of Atangana–Baleanu (ABC) fractional derivatives. By employing the next‐generation matrix method, the basic reproduction number is obtained and it is confirmed that the alcoholism‐free equilibrium is locally stable for . The Castillo–Chavez method then extends local stability to global asymptotic stability. The sensitivity analysis ranks dynamics of alcohol‐use with parameters like alcohol‐related mortality (), population growth (), transition and recovery rates (), conversion rates (), and consumption levels as the most influential factors. To illustrate the varying population characteristics, three case studies are examined, each with a different set of parameters and reflecting various physiological and behavioural risk groups. Moreover, an Artificial Neural Network (ANN) is proposed in order to not only improve the computational efficiency but also offer variable numerical forecasts. The analytical and ANN‐based numerical outcomes together provide a rich framework for the understanding and predicting of alcoholism's development.