Sensitivity analysis for variational inequalities and nonlinear complementarity problems (Q2276887)

The general parametric variational inequality problem in finite- dimensional spaces is defined as ``find \(x\in R(\epsilon)\) s.t. \(F(x,\epsilon)^ T(y-x)\geq 0\), \(\forall y\in R(\epsilon),''\) where F is a mapping from \({\mathbb{R}}^ n\times {\mathbb{R}}^ k\) to \({\mathbb{R}}^ n\), R is feasible point-to-set (set-valued) map and \(\epsilon \in {\mathbb{R}}^ k\) is the parameter vector. Its solution set is \(S(\epsilon)=\{x\in R(\epsilon)|\) \(F(x,\epsilon)^ T(y-x)\geq 0\), \(\forall y\in R(\epsilon)\}.\) If R(\(\epsilon\))\(\equiv K\), a closed convex subset of \({\mathbb{R}}^ n\), this problem is equivalent to the parametric generalized complementarity problem ``find x s.t. \(x\in K\), \(F(x,\epsilon)\in K^*\), \(F(x,\epsilon)^ Tx=0''\), where \(K^*\) is the positive dual cone of K. A special significant case is that \(K={\mathbb{R}}^ n_+.\) The subject of sensitivity analysis for variational inequalities and nonlinear complementarity problems is to investigate continuity and differentiability properties of the solution set map S as a set-valued or single-valued map. This paper surveys the main results in the area of sensitivity analysis for finite-dimensional variational inequality and nonlinear complementarity problems. It provides an overview of Lipschitz continuity and differentiability properties of S for variational inequality problems, defined on both a fixed K and general perturbed sets R(\(\epsilon\)), and for nonlinear complementarity problems. Concluding remarks and directions of future research are proposed in the last section.