Is Partial Squares Regression the Right Technique?

Using a technique called p l s r, which i believe is partial squares regression. A known technique, but not nesessarily the go to for a lot of data scientists like random forest and x g boost. The ordinary regression methods either produces coefficients with high standard errors or they usually failed completely. We decidedto use partial leaf cris gression because the main objective was not to actually predict accurately whatwas happening,. Rather understanding the importance of each of these factors such as income and population. So you are looking to predict whether or not usage would go up or down.