Comparison between use of the Partial Least Squares Regression (PLSR) and Principal Components Regression (PCR) in the factors influencing cement expansion

In this research, Two methods are used: Partial Least Squares Regression (PLSR) and Principal Components Regression (PCR) to build a model for autoclave cement on factors influnce on it. The comparison between these two methods is done by using two components for the PLSR and PCR, the plot of the fitted data shows that Partial Least Squares Regression represente the data better than Principal Components Regression, and R2 insures this result which is showen by the figure. After that, 10 variables are used to compare these methods, this comparison indicates that the two methods represente the data in the same way. The goal is inreducing the number of components used in the two methods to avoid Over- Fitting, then it is depended on cross- validation method, this method indicates that Partial Least Squares Regression method is more economic than Principal Components Regression.