Kolmogorov-Arnold neural network technique for predicting heat and mass transfer analysis of Boger nanofluid flow over a Riga surface

Boger nanofluids intensify heat transmission and elasticity control in polymer extrusion, fiber spinning, and coating processes by simulating stagnation point flow (SPF) via stretching sheet models. Real-time optimization is made possible via the Kolmogorov-Arnold networks integration, such as quick prototyping of viscoelastic coatings or predictive management of electrohydrodynamic pumps. This article provides the effects of electromagnetohydrodynamic (EMHD) and permeable medium on the SPF of Boger nanofluid via an extending sheet with heat source/sink and chemical reaction effects. The governing PDEs with boundary conditions are transformed into ODEs using suitable similarity variables. These ODEs are resolved by utilizing the Runge-Kutta-Fehlberg 4 th -5 th order (RKF-4 th 5 th ) and the shooting techniques. These engineering coefficients are scrutinized by employing the Kolmogorov-Arnold networks. Noteworthy results of the current examination are that the R 2 for the test set and validation data sets for skin friction were found to be 0.999889 and 0.999854, for Nusselt number were 0.999897 and 0.999951, and for Sherwood number were 0.999452 and 0.999509, respectively. Increasing the solvent fraction parameter causes an escalation in the velocity profile. An up surging the heat source/sink parameter improves the temperature profile. Isothermal and streamline plots have been exhibited for both the absence and presence of porous medium.