Prediction-of-cement-compressive-strength-using-stacked-ensemble-modelling
The actual concrete compressive strength (MPa) for a given mixture under aspecific age (days) was determined from laboratory. Data is in raw form (not scaled).The data has 8 quantitative input variables, and 1 quantitative output variable, and 1030 instances (observations).Context:Concrete is the most important material in civil engineering. The concrete compressive strength is a highly nonlinear function of age and ingredients. These ingredients include cement, blast furnace slag, fly ash, water, superplasticizer, coarse aggregate, and fine aggregate. Summary of steps taken and performance achieved: Multiple models with different levels of complexity were attempted. The dependent and independent variables seem to have a nonlinear relationship as the performance of models improved with increasing complexity. MAPE was selected as the evaluation metric.Regularization, feature selection and hyper-parameter tuning was employed to improve the model performance. The models attempted are Linear Regression with no regularization Ridge and Lasso Gradient Boosting Random Forest XGboost Support Vector Machine Stacking - ensemble of the best estimators of the above tuned models with a meta regressor (i.e. Ridge) which gave the best result (MAPE of less than 10)
