(2) x is an l×1 vector containing the regressor (or independent) variables. These are the variables that can be used to describe how the output will behave. In many cases, irrelevant variables may be considered in order to be certain that a complete model has been obtained. The regressors need not have any physical relationship with the output; rather, they may solely be correlated with the output variable. If they are correlated, then the quality of the model may suffer once the correlation no longer holds.