Implementation of a statistical model for measuring the compressive strength of concrete which explains 92.3% of the variance.
Img Src: GPRS
Studies have shown that the compressive strength of concrete is a highly non- linear function of the age and ingredients of the concrete. This is a model built to predict the compressive strength of concrete using its age and ingredients as regressors.
A Multivariate Adaptive Regression Spline model with Interaction terms provides an R^2 value of 87.4% whereas an ensemble random forest regression model explains 92.43% of the variance in the response variable.