A unified monotonic approach to generalized linear fractional programming (Q1422886)

The authors consider the following nonconvex optimization problems: \[ \begin{aligned} &\max \left\{ \left.\Phi \left( \frac{f_{1}(x)}{g_{1}(x)},\dots,\frac{f_{m}(x)}{g_{m}(x)}\right) \,\right|\, x\in D \right\} \qquad\text{and} \tag{P} \\ &\min \left\{ \left. \Phi \left( \frac{f_{1}(x)}{ g_{1}(x)},\dots,\frac{f_{m}(x)}{g_{m}(x)}\right) \,\right|\, x\in D \right\}, \tag{Q}\end{aligned} \] where \(D\) is a nonempty polytope in \(\mathbb{R}^{n}\), \(f_{1},\dots,f_{m}\), \( g_{1},\dots,g_{m}\) are linear affine functions on \(\mathbb{R}^{n}\) such that \[ -\infty <a_{i}:=\min_{x\in D} \frac{f_{i}(x)}{g_{i}(x)}<+\infty ,\quad i=1,\dots,m, \] while \(\Phi :\mathbb{R}^{m}\rightarrow \mathbb{R}\) is a continuous function, increasing on \( \mathbb{R}_{a+}^{m}:=\{y\in \mathbb{R}^{m}\mid y_{i}\geq a_{i}\) \((i=1,\dots,m)\},\) i.e. satisfying \[ a_{i}\leq y_{i}'\leq y_{i}\;(i=1,\dots,m)\;\Rightarrow \;\Phi (y')\leq \Phi \left( y\right) \] The authors present an efficient unified approach for solving all variants of problems (P) and (Q), based on \textit{H. Tuy} [SIAM J. Optim. 11, No. 2, 464--490 (2000; Zbl 1010.90059)].