Debrecen University, Hungary
Abstract: In this paper we discuss the theoretical backgrounds and implementation issues of scaling a linear-fractional programming problem (LFP). We consider an LFP problem in the canonical form and show how to scale rows and columns of the problem, then if the scaled problem has been solved, we show how the solution obtained may be un-scaled. We also overview briefly three scaling rules for calculating scaling factors. Finally, to illustrate how these rules work we consider a numeric example.
Keywords: linear-fractional programming, scaling
Classification (MSC2000): 90C32
Full text of the article: