Exploring Heat Transfer Enhancement: Machine Learning Predictions Using Artificial Neural Network for Water-Based Cu and CuO Micropolar Nanofluid Transportation over a Radiating Surface

Recent advancement in science is due to the innovation and invention in research with the application of machine learning approach. The artificial neural network is one of the basic methods used in machine learning. The goal of this work is to anticipate and optimize the heat transfer rates of water-based micropolar nanofluids over a radiating surface by use of artificial neural networks. The free convection impact is observed for the interaction of thermal buoyancy in the momentum equation and the insertion of the thermal radiation in the energy equation. The assumption of a suitable transformation rule involved with similarity variables is adopted to design a non-dimensional model for the proposed problem. Furthermore, the system of transformed equations is handled numerically using traditional Runge–Kutta based on the shooting technique. The analysis for the characterizing parameters is obtained graphically and described briefly. Furthermore, the increased heat transfer rates for copper–water and copper oxide–water nanofluids are estimated while accounting for a variety of affecting factors by using a machine learning technique. An artificial neural network is adopted to predict optimized heat transfer rate response for the proposed factors with error analysis. The regression analysis is presented for the choice of a suitable number of neurons with adequate testing, training, and validation data. The results of the study show that the concentration of both kinds of nanoparticles increases beyond the profiles of fluid temperature and velocity. On the other hand, CuO nanoparticles have a stronger impact on the velocity profile than Cu nanoparticles, although the temperature has the opposite effect.