Abstract Large magnetocaloric effect demonstrated by manganite-based compounds has advanced the possibility of practical implementation of magnetic refrigeration technology. In order to enhance the magnetocaloric effect of manganite based compounds and to reduce the experimental stress involved in maximum magnetic entropy change (MMEC) determination, this work develops for the first time, hybrid genetic algorithm (GA) and support vector regression (SVR) models for estimating the MMEC, using the ionic radii and crystal lattice parameters as descriptors. Aside from the performance superiority of hybrid GA-SVR-Ionic model (which uses ionic radii and the concentrations of the dopants as descriptors) as compared with GA-SVR-Lattice model, easy accessibility of its descriptors, ability to incorporate four different dopants are among the uniqueness of the proposed GA-SVR-Ionic model.