Evaluation of Compressive Strength of Self-Consolidating Concrete Using Radial Basis Function Canberra Kernel Based Support Vector Machine Model

Pertileis a glassy volcanic rock that is utilized for the application of construction requiring enhanced lightweight, acoustic, and thermal properties. A huge number of powders are gathered and stored as the material of waste during raw perlite processing. This research analyzes the powder effect of Waste Perlite (WP) on durability properties in Self-Consolidating Concrete (SCC). By utilizing different integrations of WP, Metakaolin (MK), Silica Fume (SF), and Limestone Filter (LF) are tested for 2 years by various mixtures prepared with a powder of 580 kg/m3. As per BS EN 12390–3 test approach, the compressive strength of examined mixtures is evaluated on cubic specimens. Then, the discrete density particles are evaluated by a pycnometer approach utilizing the materials of over-dry. The f_c is consistently higher for SCC in the integration of WP with SF and MK in terms of a content of cement and testing age compared to the preparation of mixture with only WP. The materials like MK, LF, SF, mixture proportion data and test results are obtained as the input data for evaluation of concrete strength. To evaluate the strength of concrete an improved Radial Basis Function Canberra kernel based Support Vector Machine model (RC-SVM) is proposed for accurate evaluation. Experimental results of proposed RC-SVM model achieved Root Mean Squared Error (RMSE) of 2.457 which is less when compared to exiting approaches like eXtreame Gradient Boosting (XGBoost).