Machine learning-based numerical analysis of ternary hybrid nanofluid flow with Cattaneo-Christov heat flux

The present study provides an analysis of the matter of ternary hybrid nanofluid flow (Al 2 O 3 –Ag–CuO/water) over stretchable surfaces with magnetic fields, thermal radiation, internal heat generation, and velocity slip, as well as Cattaneo–Christov heat flux (CCHF) using machine learning. A neural computing framework using the Levenberg–Marquardt algorithm (NC-LMA) has been developed, trained using highly accurate reference data generated using a Python-based collocation method for boundary value problems. The developed NC-LMA model was able to predict temperature and velocity profiles for varying thermophysical parameters very accurately with levels of mean square error of about 10 −14 . The performance of the proposed model was validated using regression plots, error histograms, and absolute error plots. Results showed that blade-shaped nanoparticles really enhance heat transfer. The model predicts the non-Fourier behavior of given heat conduction. Overall, the designed NC-LMA model will be a valuable and reliable tool to use for predicting the thermal characteristics of nanofluids in advanced thermal systems such as microelectronics cooling, solar collectors, and biomedical heating activities.