Machine-Learning-Based Prediction of the Compressive Strength of Superabsorbent Polymer-Incorporated Cementitious Materials: Modeling and Experimental Insights into Polymer Characteristics

This repository provides the Python code used in the study "Machine-Learning-Based Prediction of the Compressive Strength of Superabsorbent Polymer-Incorporated Cementitious Materials: Modeling and Experimental Insights into Polymer Characteristics". The code implements the M2 model for predicting the compressive strength of SAP-incorporated cementitious materials, and includes the CatBoost model (best-performing among the five ML methods evaluated) together with scripts to generate SHAP-based interpretability plots.