Numerical simulation in conjunction with artificial neural networks (ANN) has shown to be an advanced method for simulating and modeling intricate fluid dynamics problems. However, to guarantee that the model can precisely forecast transport phenomena, ANN modeling is necessary yet difficult. This study examines the symmetry of blood-based MHD squeezing nanofluid $(Au \sim Blood)$ , bi-hybrid $(Au + Fe_{3}O_{4} \sim Blood)$ , and tri-hybrid nanofluid $(Au + Fe_{3}O_{4} + MWCNTs \sim Blood)$ flow between parallel plates using a comprehensive numerical simulation and artificial neural network (ANN) analysis. The heat transfer mechanism is investigated employing the heat source/sink, thermal radiation, suction/injection, the magnetic field, and porous media. The governing partial differential equations are solved numerically with an improved finite difference approach the Keller-box technique after being modified by similarity transformations. In order to effectively predict fluid flow characteristics, this study proposes a novel approach that combines a multilayer ANN with the Levenberg–Marquardt algorithm (LMA). A strong magnetic field reduces fluid flow at the contact due to Lorentz effects, resulting in lower radial velocity as $M$ increases. In comparison to injection, the rising values of the suction parameter raise the temperature by giving the velocity of the fluid layer and eliminating isolated boundary layers. Increased permeability at the bottom plate results in higher flow resistance and reduced velocity profiles toward the upper plate. The proposed ANN approach provides fast convergence and reduced processing costs without the need for linearization. This research offers valuable insights into the performance gains made possible by tri-hybrid nanofluids, such as increased thermal conductivity, paving the way for advancements in biological applications like cancer treatment, blood pumping, and targeted drug delivery.
