Supervised machine learning analysis of Hall current and radiative effects in Casson nanofluid flow over a stretching surface with activation energy

Abstract

This study investigates the steady three-dimensional flow of a Casson nanofluid over a permeable stretching sheet, incorporating Hall current, activation energy, thermal radiation, heat generation/absorption, and a transverse magnetic field. To capture realistic heat and mass transport at low Reynolds numbers, Brownian motion and thermophoresis are also considered. The governing nonlinear PDEs are reduced via similarity transformations and solved numerically using the Runge-Kutta shooting method. Results show that the Hall parameter enhances transverse velocity by counteracting magnetic damping, while activation energy suppresses chemical reactions, leading to higher nanoparticle concentrations near the surface. Brownian motion intensifies thermal transport but decreases solutal concentration due to diffusion. A supervised machine learning model, Multiple Linear Regression (MLR), is further applied to quantify sensitivities of skin friction and heat transfer. The MLR exhibits high predictive accuracy (R2 > 0.97; MSE < 2.7×10−4), confirming robustness. Parametric analysis identifies suction velocity (S), thermophoresis (Nt), and magnetic field strength (M) as the most influential factors controlling drag and heat transfer. The findings agree closely with existing literature, validating the approach. This integrated numerical-statistical framework offers practical insights for optimizing nanofluid-based systems in energy harvesting, thermal management, and MHD applications. © 2025 Informa UK Limited, trading as Taylor & Francis Group.