Failure Mode Detection of Reinforced Concrete Shear Walls Using Ensemble Deep Neural Networks

Abstract Reinforced concrete structural walls (RCSWs) are one of the most efficient lateral force-resisting systems used in buildings, providing sufficient strength, stiffness, and deformation capacities to withstand forces generated during earthquake ground motions. Identifying failure mode RCSWs is a critical task that can assist engineers designers choosing appropriate retrofitting solutions. This study evaluates efficiency three ensemble deep neural network models, including model averaging ensemble, weighted average integrated stacking for predicting RCSWs. The models compared against previous studies traditional well-known (AdaBoost, XGBoost, LightGBM, CatBoost) machine learning methods (Naïve Bayes, K-Nearest Neighbors, Decision Tree, Random Forest). proposed as best-suited prediction identifying since it has highest accuracy, precision, recall among alternative models. In addition, complex advanced learning-based commonly referred black-box, SHapley Additive exPlanation method also interpret workflow illustrate importance contribution components impact determining