ABSTRACT
In statistical optimization, the function to be optimized is called the objective function, which is an explicit mathematical function that describes what is considered the optimal solution. There is a mathematical model that relates a random variable, called the criterion or dependent variable, to a vector of unknowns, called regression coefficients, and a vector of predictor variables, which are sometimes called independent variables. The objective of regression is to evaluate the coefficients of an equation relating the criterion variable to one or more other variables, which are called the predictor variables. Regression is a means of calibrating the unknown coefficients of a prediction equation, whereas correlation provides a measure of goodness of fit. Since most computer programs that perform multiple regression analyses include goodness-of-fit statistics as part of the output, it is important to recognize the meaning of these statistics.
