Convex functions and their applications. A contemporary approach (Q5916039): Difference between revisions

This book is the second edition of [Zbl 1100.26002], containing about 160 pages of new material as well as some corrections and updates in comparison with the original one. It is divided into seven chapters preceded by a list of symbols and followed by five appendices and the usual lists of references (limited to exactly 500 entries) and indices. The new material can be found in Chapters 2--7 (entirely new are Chapters 2 and 4--6) and in Appendix B. This 415 pages long book is an excellent resource for anyone interested in convex functions and (some of) their applications, taking the reader in a trip over territories that lie way beyond the classical theory. One can note an emphasis on inequalities, however they are not the main subject of the volume, this role being covered by convexity in several of its incarnations. The first chapter is dedicated to real-valued convex functions on intervals, presenting besides their definition and the ones of various objects related to them such as conjugate function subdifferential etc. also the most important results involving conjugate functions such as discrete and integral Jensen-type inequalities, the Young inequality, the Hermite-Hadamard inequality or characterizations of convex functions and their topological properties. In the second chapter, one can read about convex sets in real linear spaces, where, besides the necessary definitions and the most important properties of convex sets, classical themes like orthogonal projections, hyperplanes, separation, ordered linear spaces and symmetric real matrices as well as some rather special issues such as Clarkson's inequalities are treated. The longest chapter in the book is the third one whose contents evolves around convex functions on normed linear spaces. After the classical definitions and some illustrative examples, the connections between convex functions and convex sets are presented. Then, one can read about (sub)differentiablity properties of convex functions, positive homogeneous convex functions and extrema of convex functions, and also about not so ``mainstream'' things like some inequalities associated to perspective functions or the Prékopa-Leindler inequality. The next chapter is devoted to the connections between convexity and the theory of majorization, the classical results due to Hardy-Littlewood-Pólya, Schur and Horn, among others, being listed. Moreover, one can find some eigenvalue inequalities and some things about the theory of hyperbolic polynomials as well as about vector majorization in Euclidean spaces. Chapter 5 presents various facts about convexity in matrix spaces such as matrix convexity, convex spectral functions and even geodesic convexity in the space of positively definite matrices endowed with the trace metric. It is followed by a chapter on duality and convex optimization, where not only the classical Legendre-Fenchel duality theory is presented but also different facts about the correspondence of properties under duality, minimax inequalities, Moreau-Yosida approximation and even the Hopf-Lax formula for the Hamilton-Jacobi equation. Chapter 7 is a new and enhanced version of Chapter 4 from the first edition of the book and deals with new developments in majorization theory, more precisely with Choquet theory and its extension to a special class of signed measures called Steffensen-Popoviciu measures. The five appendices present some additional results to the main chapters. The first of them is devoted to generalized convexity on intervals, containing a short version of Chapter 2 from the first edition of the book. The next one contains some background material on convex sets, from the Hahn-Banach extension theorem to the Krein-Milman theorem. In Appendix C, some elementary symmetric functions are presented, with emphasis on Newton's inequalities and symmetric polynomial majorization. Next, one can read about second-order differentiability of convex functions, more precisely about Rademacher's and Alexandrov's theorems. Last but not least in Appendix E, basic facts about a variational approach of the partially differential equations by means of convexity are presented, including some applications to elliptic boundary-value problems and the celebrated Galerkin method. The sections in the book end with several exercises and the chapters with interesting comments about their contents and beyond, even surprising connections to other branches of mathematics being mentioned here and there. Note also that some of the results presented in the book are new in the literature. This book can prove to be useful for both researchers and instructors (e.g. as a graduate coursebook) as the material is clearly presented and most of it can be understood without very deep knowledge of convex analysis or functional analysis. Finally, I can only second the reviewer of the first edition of this book who wrote ``the book will be useful to all who are interested in convex functions and their applications.''