Finite dimensional convexity and optimization (Q5956951)

The main aim of the book is to present notions of convex analysis in a finite dimensional setting. The book is divided into 7 chapters. The first three of them are devoted to the notions and main results of convexity in \(\mathbb R^n\). The topic of the fourth Chapter is Linear Programming. The analysis of convex functions starts with Chapter 5. In Chapter 6 the differentiability properties of convex functions are studied. Finally Chapters 7 and 8 deal with constrained convex and also nonconvex optimization problems. Since the book is intended to be useful for a broad audience the authors have restricted the material to the finite dimensional case. The proofs in the book are written with a high level of rigor. The exercises at the end of the chapters give nice complements and applications. The results contained in the book are nowadays well-known. The first chapter (Convexity in \(\mathbb R^n\)) introduces the definition and main topological properties of convex sets in \(\mathbb R^n\). The second chapter (Separations and Polarity) deals with separation theorems in \(\mathbb R^n\). Here the authors also present the projection map and the Minkowski-Farkas Lemma. The third chapter (Extremal Structure of Convex Sets) deals with geometric properties of convex sets in \(\mathbb R^n\). The theorems of Krein-Milman and of Weyl are given and the relationship between convex sets, and their extreme points are extensively discussed. Chapter 4 (Linear Programming) gives a brief introduction to linear programming with emphasis on optimality conditions and duality theory. The only algorithm presented is the simplex methods. The Minkowski-Farkas Lemma is used to provide optimality conditions and for establishing the duality theory. In the Chapter 5 (Convex Functions) some important properties of convex functions are given. In particular the continuity of such functions on the interior of its domain is proved. Furthermore it is proved that any pointwise limit of a sequence of convex functions defined on a relatively open set \(U\) is convex, continuous, and is the uniform limit on any compact subset of \(U\). The sixth chapter (Differential Theory of Convex Functions) presents three notions of differentiability for convex functions. The first one is the directional derivative, the second one the subgradient and the third one the well-known differentiability of a function. Furthermore it is proved that subdifferentials of a pointwise convergent sequence of convex functions (like those studied in chapter five) is also pointwise convergent. The Hahn-Banach Theorem is presented and used in this chapter to prove the above result. The seventh chapter (Convex Optimization with Convex Constraints) deals with constrained convex optimization problems. On the one hand sufficient conditions are given for the existence of solutions. On the other hand the well-known criteria of Kuhn-Tucker are presented. Finally in the last Chapter 8 (Nonconvex Optimization) the analysis of optimization problems is extended to two more cases. In the first section necessary and sufficient conditions for a point to be a solution to a minimization problem with a quasi-convex objective function and convex constraints are given. In a second section the optimality conditions for optimization problems with continuously differentiable, but not necessarily convex data are studied. The book is appropriate as a class text, or also suitable for self-study.