A NEW HYBRID FRAMEWORK OF MACHINE LEARNING TECHNIQUE IS USED TO MODEL THE COMPRESSIVE STRENGTH OF ULTRA-HIGH-PERFORMANCE CONCRETE

To calculate the compressive strength (CS) of concrete, it is necessary to investigate Ultra -High-Performance Concrete (UHPC) in terms of its components and their quantities. Empirical analysis of relationships between constituents can be more time-and money-consuming. The CS can now be evaluated based on the composition of the ingredients thanks to intelligent systems. Additionally, it is advisable to promote the use of eco-friendly materials in concrete, one of the most commonly used materials in the world. The CS of UHPC was attempted to model in this study. The CS of concrete has been simulated using Support Vector Regression (SVR), a Machine Learning (ML) technique that is compatible with Particle Swarm Optimisation (PSO) and Henry's Gas Solubility Optimisation (HGSO), based on various materials used in the construction present article. The CS values were determined through the testing of eight components. The modeling process was evaluated using a variety of metrics. In this regard, the test phase modeling's root-mean-square error (RMSE) for SVR - HGSO was 8.45, while it was 9.23 for SVR - PSO. SVR - HGSO ' s RMSE rate for the training phase was calculated at 10.15, which is 3.3 percent higher than SVR - PSO ' s RMSE of 10.49.