Machine Learning Techniques Applied to Uniaxial Compressive Strength of Oporto Granite

摘要

This article employs two machine learning techniques, viz., Least Square Support Vector Machine (LSSVM) and Multivariate Adaptive Regression Spline (MARS), for determination of Uniaxial Compressive Strength (σ_c) of oporto granite. LSSVM uses a quadratic cost function. MARS is a nonparametric regression technique. Free porosity (N_(48)), dry bulk density (d) and ultrasonic velocity (v) have been used as input of the LSSVM and MARS models. The output of LSSVM and MARS is σ_c. The developed LSSVM and MARS give equations for prediction of σ_c. A comparative study has been carried out between the developed LSSVM, MARS, Support Vector Machine (SVM) and Artificial Neural Network (ANN) models. The results show that the developed LSSVM and MARS models are efficient tools for determination of σ_c of Oporto granite.