This research explores the characteristics of electrode coatings intended for application in weld joints within nuclear power plants. Specifically, the study focuses on evaluating the electrical, thermophysical, and physicochemical properties of shielded metal arc welding electrodes formulated with an Al2O3-CaF2-CaO-SrO-based composition. These electrodes were developed using an extreme vertices design technique. To assess the coating's structure and phases, X-ray diffraction analysis was employed, while Fourier transform infrared spectroscopy was utilized to identify the types of chemical bonds within the composition. Advanced characterization techniques were applied to analyze the physicochemical, thermophysical, and structural attributes of the coating formulations. Thermal properties such as specific heat and conductivity were examined using a hot disk apparatus. Additionally, thermogravimetric analysis was utilized to ascertain the enthalpy change and thermal stability of the flux coating. Statistical analysis was applied to construct regression models for various coating properties, aiming to explore how the mineral constituents influence these properties within the flux coating. Regression analysis is a statistical tool employed to establish connections between mineral interactions. Furthermore, artificial neural network (ANN) models were constructed and compared with regression analysis to evaluate prediction accuracy. In the context of selecting flux compositions, it aids in understanding the meaningfulness of each factor and its connection to the coating's overall performance. The findings reveal that the individual elements of mineral constituents, as well as their binary and tertiary interactions, have a noteworthy impact on the physicochemical, electrical, and thermophysical properties of the flux composition.