Concrete Compressive Strength Prediction Based Web Application Using Gradient Boosting Based ML Ensemble: A Case Study

The escalating demand of high strength concrete for the construction of the high-rise buildings is straining the cement industry. The experimental data for calculating compressive strength (MPa) is produced at the laboratory for various sets of trial batches of concrete under a specific age with ingredients like cement, blast furnace slag, fly ash, water, superplasticizer, coarse aggregate, and fine aggregate. The raw data is pre-processed and used to train the state-of-the-art machine learning models for predicting the concrete compressive strength. Gradient boosting based ensemble using blending and stacking is implemented in order to optimize the predictive performance. However, despite of that, a fine-tuned Cat Boost model, using Random and Grid search parameters, performed the best, and was selected as the proposed model.