<![CDATA[Clay bricks are a very popular building material used in projects in Indonesia, with compressive strength as a key quality indicator. Laboratory compressive strength testing is destructive, time-consuming, and less practical for small and medium-scale industries. The study seeks to create a compressive strength predictive model utilizing production statistics and advanced machine learning algorithms and to compare manually molded and pressed bricks. The dataset consists of secondary production data, including clay composition, moisture content, molding pressure, firing temperature, and firing time, with compressive strength as the output variable. Three algorithms were applied: Linear Regression, Support Vector Regression, and Random Forest Regression. To assess model performance, utilized the coefficient of determination (R²), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Study outcomes reveal that Random Forest Regression achieves the best performance, with an R² value of 0.89 and the lowest prediction errors. Results from the feature importance analysis demonstrate that molding pressure and firing time are the most influential factors affecting compressive strength. The predicted results are consistent with previous experimental studies reporting higher and more stable compressive strength in pressed bricks compared to manually molded bricks. This approach demonstrates strong potential as a data-driven decision-support tool for brick quality control.]]>
Copyrights © 2026
