Minimization of the sum of several linear fractional functions

From MaRDI portal





The authors consider a rank-\(p\) linear-fractional programming problem, which consists in minimizing the sum of \(p\) ratios of affine functions over a polytope (the denominators are assumed to be strictly positive on the feasible set). By means of the Charnes-Cooper transformation, this problem can be reduced to an equivalent problem in which the objective function is the sum of \(p-1\) ratios with a linear function, for which primal-dual parametric simplex algorithms are proposed if \(p= 2\) or \(3\). The new problem can in turn be reformulated, after introducing \(2p-2\) auxiliary variables, into a concave minimization problem on a convex set, the objective function of which is the optimal value function of a convex minimization problem; a generalized convex multiplicative programming algorithm for solving this problem is discussed. Finally, another reformulation leads to a linear programming problem with \(p-1\) additional nonconvex constraints of a simple type; a branch-and-bound algorithm for solving this problem is presented. Computational experience showing that up to rank-10 problems can be solved by these methods is reported. For the entire collection see [Zbl 0961.00019].




Cited in
(22)








This page was built for publication: Minimization of the sum of several linear fractional functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2751321)