Unifying local-global type properties in vector optimization (Q1756763)

After an introduction, Section 2 presents some preliminaries. In particular, the new concept of algebraic local \(K\)-extremality for vector functions with respect to a set \(K\) is introduced. Moreover, some important classes of generalized convex scalar and vector functions which are known to possess so-called `local min-global min' and/or `local max-global min'-type properties in scalar, unified vector and vector optimization with respect to a convex cone are reviewed. Section 3 extends some notions and introduces new concepts of generalized convexity for vector-valued functions with respect to two arbitrary sets \(K_1\) and \(K_2\) of the outcome space. In particular, the semi-strict and explicit \((K_1, K_2)\)-quasi-convexity (Section 3.1), the unidirectional \((K_1, K_2)\)-quasi-convexity (Section 3.2) and that for the special case of \(K = K_1 = - K_2\) (Section 3.3) as well as the bidirectional \((K_1, K_2)\)-quasi-convexity (Section 3.4) are defined. Some relationships between these new concepts are presented. Section 4 deals with some local-global extremality properties and presents the main results of this paper. Theorems 9--11 present general local-global extremality principles for bidirectional \((K_1,K_2)\)-quasi-convex functions, for unidirectional \((K_1,K_2)\)-quasi-convex functions and for semi-strictly \((K_1,K_2)\)-quasi-convex functions, respectively. In Theorems 12--15, the authors derive `local min-global min' and `local max-global min'-type properties, respectively. The paper finishes with some concluding remarks and points out some further extensions of this work, e.g., to establish similar local-global extremality properties for set-valued functions with respect to variable ordering structures.