Experience gained during a ten-year long involvement in modelling, program ming and application in nonlinear optimization helped me to arrive at the conclusion that in the interest of having successful applications and efficient software production, knowing the structure of the problem to be solved is in dispensable. This is the reason why I have chosen the field in question as the sphere of my research. Since in applications, mainly from among the nonconvex optimization models, the differentiable ones proved to be the most efficient in modelling, especially in solving them with computers, I started to deal with the structure of smooth optimization problems. The book, which is a result of more than a decade of research, can be equally useful for researchers and stu dents showing interest in the domain, since the elementary notions necessary for understanding the book constitute a part of the university curriculum. I in tended dealing with the key questions of optimization theory, which endeavour, obviously, cannot bear all the marks of completeness. What I consider the most crucial point is the uniform, differential geometric treatment of various questions, which provides the reader with opportunities for learning the structure in the wide range, within optimization problems. I am grateful to my family for affording me tranquil, productive circumstances. I express my gratitude to F.

Klappentext

This book is the first uniform, differential geometric approach to smooth nonlinear optimization. This advance allows the author to improve the sufficiency part of the Lagrange multiplier rule introduced in 1788 and to solve Fenchel's problem of level sets (1953) in the smooth case. Furthermore, this permits the author to replace convexity by geodesic convexity and apply it in complementarity systems, to study the nonlinear coordinate representations of smooth optimization problems, to describe the structure by tensors, to introduce a general framework for variable metric methods containing many basic nonlinear optimization algorithms, and - last but not least - to generate a class of polynomial interior point algorithms for linear optimization by a subclass of Riemannian metrics. Audience: The book is addressed to graduate students and researchers. The elementary notions necessary for understanding the material constitute part of the standard university curriculum.

Inhalt
Preface. 1. Introduction. 2. Nonlinear Optimization Problems. 3. Optimality Conditions. 4. Geometric Background of Optimality Conditions. 5. Deduction of the Classical Optimality Conditions in Nonlinear Optimization. 6. Geodesic Convex Functions. 7. On the Connectedness of the Solution Set to Complementarity Systems. 8. Nonlinear Coordinate Representations. 9. Tensors in Optimization. 10. Geodesic Convexity on R°n+ 11. Variable Metric Methods Along Geodesics. 12. Polynomial Variable Metric Methods for Linear Optimization. 13. Special Function Classes. 14. Fenchel's Unsolved Problem of Level Sets. 15. An Improvement of the Lagrange Multiplier Rule for Smooth Optimization Problems. A. On the Connection Between Mechanical Force Equilibrium and Nonlinear Optimization. B. Topology. C. Riemannian Geometry. References. Author Index. Subject Index. Notations.