Punching strength in Slab-Column Connections includes Shear Reinforcement using optimized tree schemes

The current work introduces a data-driven model to detect punching strength (Vn) in slab-column junctions with shear reinforcement. For the prediction aim, the data collection was created including nine input parameters based on the punching shear mechanism from literature. This data collection comprises 327 rows of samples. The learning collection (70%), validating collection (15%), and evaluation collection (15%) of the dataset were utilized in particular in the building, validating, and testing stages of the suggested framework.  This exploration considered a decision tree (DT) algorithm linked with two hybrid optimized methods, named Manta ray foraging optimization (MRFO) and Victoria Amazonica optimization (VAO), as an optimizer. After further inspection, the VAO-DT scheme recommended in this exploration generated superior outcomes to those of earlier studies included in the current study. Based on the presented literature, the greater R2 and lower RMSE and MAE values indicate that the VAO-DT model produces more robust and dependable outcomes than extreme gradient boosting (XGB) and hybrid form of it. MRFO-DT demonstrated good procedural reliability with R2 quantities of 0.9826, 0.9893, and 0.9858 during training, validation, and appraisal. VAO-DT surpassed MRFO-DT in R2 quantities, with 0.9943, 0.9988, and 0.9954. Although MRFO-DT could receive an acceptable function, the DT model linked with VAO can be recognized as the most potent model for forecasting purposes and can be utilized in practical applications.