Optimizing Concrete Strength Prediction: A Random Forest Approach with Hyperparameter Tuning through Random Search for Enhanced Construction Quality

The goal of this study is to help with the creation of concrete mixes and the building of structures, so that construction projects last longer and are more stable. As the building industry changes, it’s important to use advanced machine learning methods to get the best predictions for concrete strength and better total construction quality. During our tests, we checked how well the model worked using set parameters, Our main goal was to get a Mean Absolute Error (MAE) of less than 4 and a $R^{2}$ value of above 90%, which means that our forecasts were very accurate and reliable. We found that our model regularly did better than these benchmarks in a variety of testing conditions by doing a lot of testing. The mean absolute error (MAE) for our model was much less than 4 and the dependability ($\mathrm{R}^{2}$) was over 90%, showing that it was very good at making predictions and very reliable. Also, the RMSE values showed that the model was very good at guessing the strength of concrete with little difference from the actual values. The outcomes show that the Random Forest method works well at going above and beyond set performance goals.