Enhancing the mechanical properties’ performances coconut fiber and CDW composite in paver block: multiple AI techniques with a Performance analysis
G. Uday Kiran, G. Nakkeeran, Dipankar Roy, Sumant Nivarutti Shinde, George Uwadiegwu Alaneme
Abstract
The present research incorporates five AI methods to enhance and forecast the characteristics of building envelopes. In this study, Response Surface Methodology (RSM), Support Vector Machine (SVM), Gradient Boosting (GB), Artificial Neural Networks (ANN), and Random Forest (RF) machine learning method for optimization and predicting the mechanical properties of natural fiber addition incorporated with construction and demolition waste (CDW) as replacement of Fine Aggregate in Paver blocks. In this study, factors considered were cement content, natural fine aggregate, CDW, and coconut fibre, while the resulting measure was the machinal properties of the paver blocks. Furthermore, machine learning techniques to precision the predicting machinal properties were extensively evaluated. The outcomes from both the training and testing phases demonstrated the strong predictive power of RSM, SVM, GB, ANN, and RF with a criterion used Root Mean square error (RMSE), Mean square error (MSE), Mean Absolute Error (MAE) and correlation coefficient (R). Moreover, the results demonstrated that GB and ANN provide enhanced performance in comparison to SVM and RF for determining testing factors.