Least Squares and Entropy: A Penalty Function Perspective

Mathematical measures of entropy as defined by Shannon and cross entropy as defined by Kullback and Leibler are currently in vogue in the field of econometrics, primarily due to the comprehensive work of Golan, Judge, and Miller. An alternative interpretation of the entropy measure as a penalty function over deviations is presented, and a number of parallels are drawn with least squares estimators. It is demonstrated that both approaches may be applied to the general linear model. The causes of differences in estimated parameter values are described, and some suggestions for the formulation of entropy-based econometric problems are presented.