Soft computing approaches for mechanical property predictions for polypropylene fibre in Fly Ash Mortar based machine learning

Abstract

The current study combines four techniques: multi-linear regression (MLR), artificial neural networks (ANN), support vector machine (SVM) and Random Forest (RF) to introduce a novel, alternative approach to predict compressive strength using artificial intelligence techniques and modulus of elasticity of polypropylene fibre Mortar mixed with fly ash. Inputs included cement content, Fly Ash, and polypropylene fibre; the output was mortar compressive strength and modulus of elasticity. The four methods were compared according to their accuracy and stability to predict compressive strength. The results from training and testing models have shown the great potential of MLR, ANN, SVM and Random forest in predicting the compressive strengths and modulus of elasticity of polypropylene fibre mortar. Further, the study demonstrated that SVM and ANN are preferable to MLR and Random forest when estimating experimental parameters.