Entropy optimization analysis in radiative flow of rheological material using artificial neural networks

• Implications of artificial neural networks analysis for hydromagnetic Reiner-Rivlin material is addressed. • Artificial neural networks is suggested to discuss characteristics of rheological material flows like analysis, testing, authentication and training. • Entropy generation in presence of heat generation/absorption, viscous dissipation and Ohmic heating effects is addressed. This study explores the magnetohydrodynamic dissipative flow of rheological material by stretched cylinder. Furthermore, the advanced computational method of neural networks constructed Levenberg Marquardt technique provide outstanding skills in precisely obtaining the sophisticated solutions of solutal and thermal transfer rates in highly nonlinear fluid flow problems. In the field of artificial neural networks approach the Levenberg-Marquardt technique is identified through its innovative stability and provides computational outcomes of the Reiner-Rivlin material flow employing mean square errors, validation check, error histogram, regression plots, fitness curve and comparison solution. Entropy rate in presence of heat generation, dissipation and Joule heating is deliberated. Chemical reaction of first order is taken. Ohmic heating, heat generation/absorption and dissipation are considered in heat equation. Nonlinear ordinary expressions are obtained through employing suitable transformations. The dimensionless ordinary differential expressions are solved employing bvp4c via MATLAB and then advanced computational technique of artificial neural networks is implemented to train the given datasets to increase predictive capabilities for advanced solutions. Physical description for rate of entropy, temperature, liquid flow and concentration are examined. The designed approach covers a series of actions depending on training, authentication and testing by employing a given datasets for different flow problems components. Furthermore, the comparison of artificial neural networks algorithm and bvp4c method is discussed. Clearly one can find that higher magnetic field leads to velocity reduction. Higher estimation of heat generation parameter correspond to intensify the thermal field. Decreasing trend for concentration for larger Schmidt number is detected. An intensification in entropy rate through larger Brinkman number is witnessed.