Solar collectors play a crucial role in harnessing solar radiation and converting it into thermal energy, functioning as efficient heat exchangers. Among them, solar dish concentrators are particularly notable for their ability to operate at high temperatures, making them an effective solution for both heat and electricity generation. Owing to their high efficiency in capturing and utilizing solar energy, dish collectors have attracted significant interest in solar thermal applications. These concentrators come in various cavity receiver designs—such as open, spiral, hollow, and volume configurations—allowing for versatile energy conversion. Building on this concept, the present study investigates natural convection heat transfer within a two-dimensional ‘C’-shaped cavity filled with a porous medium and hybrid nanofluids, specifically Ag-MgO (silver-magnesium oxide) and Ag-TiO \(_2\) (silver-titanium dioxide oxide). The cavity features adiabatic upper and lower surfaces, with a heated slit on the left and a cooled wall on the right. As solar devices become more compact and efficient, the shape of the cavity plays a critical role in ensuring proper thermal management to prevent overheating and sustain optimal performance. To enhance heat transfer in solar collectors, the study applies a machine learning technique, evaluating the influence of two distinct hybrid nanoparticles. Furthermore, machine learning is used to analyze how different parameters vary with the type of nanoparticle, aiming to determine the most effective combination for optimizing heat transfer. The governing equations are solved using the finite difference method coupled with the Marker and Cell (MAC) technique. The findings indicate that an increase in the Rayleigh number improves heat transfer owing to intensified buoyancy-driven convection, with Ag-MgO exhibiting greater efficacy compared to Ag-TiO \(_2\) . Raising the nanoparticle volume fraction significantly boosts heat transfer at \(\textrm{Ra}=10^6\) , with Ag-MgO and Ag-TiO \(_2\) nanofluids showing improvements of 12.32% and 11.93%, respectively. ANN analysis identifies Darcy number, Rayleigh number, and nanoparticle volume fraction as primary influencers of Nusselt number. For Ag-MgO, their impacts are 37.15%, 22.15%, and 13.79%, while Ag-TiO \(_2\) shows similar contributions: 37.07%, 23.51%, and 13.79%. At 5% volume fraction, Ag-MgO outperforms Ag-TiO \(_2\) by 11.35% at \(\textrm{Ra}=10^5\) and maintains a 0.451% lead at \(\textrm{Ra}=10^6\) , indicating consistently superior thermal performance.
