Paper Title
A Study on Performance Improvement Method of Machine Learning Model to Predict Demolition Waste Generation

Abstract
This study aimed to develop machine learning models to predict demolition waste generation (DWG) and improve their prediction performance through ensemble models and hybrid models with applying principal component analysis (PCA). Through this study, we developed machine learning models with improved prediction performance. We found that the hybrid models with PCA were effective in improving the prediction performance of DWG. However, the effect of PCA was found to be more significant when applied to the KNN algorithm than the DT and LR algorithms used in this study. Therefore, in the future, developing models that can improve the prediction performance of DWG by applying a wider range of machine learning algorithms will be necessary. Keywords - Demolition waste, Waste generation, Machine learning, Ensemble model, Hybrid model, Principal component analysis.