Constrained regression models for optimization and forecasting
AbstractLinear regression models and the interpretation of such models are investigated. In practice problems often arise with the interpretation and use of a given regression model in spite of the fact that researchers may be quite "satisfied" with the model. In this article methods are proposed which overcome these problems. This is achieved by constructing a model where the "area of experience" of the researcher is taken into account. This area of experience is represented as a convex hull of available data points. With the aid of a linear programming model it is shown how conclusions can be formed in a practical way regarding aspects such as optimal levels of decision variables and forecasting.
The following license applies:
Attribution CC BY
This license lets others distribute, remix, tweak, and build upon your work, even commercially, as long as they credit you for the original creation.