PREDICTION OF DETERIORATION COMPONENTS OFSTEEL BEAMS USING MACHINE LEARNING METHODS

The evaluation of the collapse of structural systems under seismic loading necessitates theidentification and quantification of deterioration components (DCs). In the case of Steel W-sectionbeams (SWSB), three distinct types of DCs have been derived. These deterioration components forsteel beams comprise the following: Pre-capping plastic rotation (Ɵp), post-capping plastic rotation(Ɵpc), and cumulative rotation capacity (Λ). The primary objective of this research is to employ amachine learning (ML) model for the accurate determination of deterioration components. AdaBoostand XGBoost algorithms have been used to predict deterioration components. The XGBoost has highperformance and accuracy. The evaluation metrics of the XGBoost model are as follows: R۲=۰.۸۹ andRMSE=۰.۰۰۴۶ for Ɵp, R۲=۰.۹, and RMSE=۰.۰۲۴ for Ɵpc and R۲=۰.۸۹ and RMSE=۰.۲۰۸ for Λ.

کلیدواژه ها:

Xgboost ، Adaboost ، Machine Learning ، Deterioration Components ، Steel Beams