Entropy optimization and mathematical programming (Q1374472)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Entropy optimization and mathematical programming |
scientific article |
Statements
Entropy optimization and mathematical programming (English)
0 references
8 December 1997
0 references
The book is concerned with theoretical or computational aspects of mathematical programming models in \(R^n\), generally linear or quadratic, having as objective or added to the objective and/or to the constraints the function \(\Sigma_ix_i\ln x_i\), known under the name of entropy function, measuring the amount of energy or information or uncertainty into a system. The book is divided into seven chapters. Chapter 1 introduces the concept of entropy and entropy optimization principles. Chapter 2 reports some important practical applications of entropy optimization in a number of selected fields, such as Queueing Theory, Transportation Planning, Input-Output Analysis, Regional Planning, Portofolio Optimization, Image Reconstruction. Chapter 3 deals with methods for solving entropy optimization problems subject to linear constraints, while Chapter 4 handles problems with quadratic, entropic, and general convex constraints. One of the authors' main contributions, an unconstrained dual-perturbation approach for solving linear, quadratic or convex entropy optimization models is discussed in detail in both chapters in addition to reviewing other major existing methods. Based on results obtained in Chapter 3, Chapter 5 introduces the entropic perturbation approach, i.e. the objective is perturbated by using the entropic function, for solving four classes of problems, namely, linear programming problems in Karmarkar's form, linear programming programs in standard form convex quadratic and linear and convex quadratic semi-infinite programming problems. Chapter 6 is dedicated to the \(Lp\)-Norm Perturbation Approach, a generalization of entropic perturbation. A relation between entropy optimization and Bayesian statistics, as well as some other related results, such as entropic regularization for min-max or semi-infinite min-max problems, are discussed in Chapter 7. Although after reading and then after thinking quite a lot on the book, the reviewer is still in an unusual state of mystery and not very well convinced, in fact on the Clausius or Shannon original idea also, why he should choose the entropic approach for modeling such a great variety of real life problems or on the superiority or robustness in computing solutions to them, without any doubt he can say that the book is a good and interesting one for all reading researchers, especially for those with a good philosophical feeling. A hard working researcher in the field of computational optimization can also find a lot of ideas to test for efficiency in problem solving.
0 references
entropic penalty
0 references
interior point methods
0 references
entropy function
0 references
entropy optimization
0 references
entropic perturbation
0 references
Bayesian statistics
0 references
entropic regularization
0 references
min-max problems
0 references