The production of construction waste during the renovation of schools causes significant economic and environmental challenges, highlighting the need for a reliable tool to estimate waste costs during the planning phase. Based on project-level variables available during renovation planning, a Random Forest regression model was developed to predict the total construction waste cost of public-school renovation projects. The model used ten derived material-waste ratios together with renovation area as project-level predictors, while the response variable was the total direct construction waste cost. The model was developed using data from 50 completed school renovation projects in Iraq. Its predictive performance was evaluated against Linear Regression, Decision Tree Regression, and Support Vector Regression models using the same training and testing datasets and 5-fold cross-validation. Feature-importance analysis was used to identify the predictors contributing most strongly to the predicted waste cost. Among the evaluated models, Random Forest achieved the best predictive performance, with a Mean Absolute Percentage Error (MAPE) of 11.14% and an R² of 0.944. Brick waste ratio and renovation area were identified as the most influential predictors of waste cost. The proposed Random Forest model provides a practical decision-support tool for estimating construction waste costs during the planning phase of school renovation projects, supporting more efficient budgeting and sustainable construction waste management. The findings are applicable within the scope of the investigated dataset and provide a basis for future validation using larger and more diverse project datasets.
