Determination of the Effect of XGBoost’s Parameters on a Structural Problem

Retaining walls are special engineering structures widely used in civil engineering applications for different purposes. Reinforced concrete retaining walls have both structural constraints and constraints such as overturning, shear and soil-bearing capacity. In this chapter, a dataset is generated by optimizing the cantilever-type reinforced concrete retaining wall with Teaching Learning Based Optimization (TLBO). This dataset is analyzed with Extreme Gradient Boosting (XGBoost), one of the machine learning models, and the effect of model parameters on the success of the model is investigated. Performance evaluation was performed using coefficient of determination (R2). As a result, the “colsample_bytree” parameter had the highest impact on the model.