Two-dimensional Reynolds-Averaged Navier–Stokes simulations are performed to study a single, round impinging jet heat transfer problem, utilizing the generalized k-omega (GEKO) turbulence model as a benchmark. The simulations are performed at a jet Reynolds number of 23,300 and a nozzle-to-plate distance of 2.0 where a second peak in surface Nusselt number is observed. The effects of the three primary (Csep, Cmix and Cnw) and three auxiliary (Cbf,l, Cbf,t and Cnw,sub) GEKO calibration parameters are investigated. The results indicate that Cmix has the most significant impact on the laminar-turbulent transition zone. A deep learning based regression model is developed and trained using the simulation outputs for fast predictions of the heat transfer curve. Using Csep=1.1, Cmix=−0.7, Cnw=2.0, Cnw,sub=2.25 and Cbf,t=3.0 along with laminar-to-turbulent transitional modeling values, provides the best agreement with experimental results from previous studies.