Skip to main content

1999 | Buch

Multiobjective Heuristic Search

An Introduction to intelligent Search Methods for Multicriteria Optimization

verfasst von: Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar

Verlag: Vieweg+Teubner Verlag

Buchreihe : Computational Intelligence

insite
SUCHEN

Über dieses Buch

A large number of problems require the optimization of multiple criteria. These crite­ ria are often non-commensurate and sometimes conflicting in nature making the task of optimization more difficult. In such problems, the task of creating a combined opti­ mization function is often not easy. Moreover, the decision procedure can be affected by the sensitivity of the solution space, and the trade-off is often non-linear. In real life we traditionally handle such problems by suggesting not one, but several non-dominated solutions. Finding a set of non-dominated solutions is also useful in multistaged opti­ mization problems, where the solution of one stage of optimization is passed on to the next stage. One classic example is that of circuit design, where high-level synthesis, logic synthesis and layout synthesis comprise important stages of optimization of the circuit. Passing a set of non-dominated partial solutions from one stage to the next typically ensures better global optimization. This book presents a new approach to multi-criteria optimization based on heuristic search techniques. Classical multicriteria optimization techniques rely on single criteria optimization algorithms, and hence we are either required to optimize one criterion at a time (under constraints on the others), or we are asked for a single scalar combined optimization function. On the other hand, the multiobjective search approach maps each optimization criterion onto a distinct dimension of a vector valued cost structure.

Inhaltsverzeichnis

Frontmatter
Chapter 1. Introduction
Abstract
In the past three decades, researchers in the area of heuristic search have focussed on the central theme of knowledge versus search in some form or the other. The major attention has been towards quantifying problem specific knowledge in terms of a heuristic evaluation function and developing algorithmic frameworks for solving problems in an efficient manner. The nature of the heuristic function and its effect on the search mechanism has been a subject of considerable interest.
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 2. The Multiobjective Search Model
Abstract
In the course of applying heuristic search techniques to practical problems, several issues have assumed significance. Some of these issues have evolved from practical considerations such as space and time constraints, that are directly related to the feasibility of the existing search techniques. Other issues have gained importance through the requirements of frameworks for modeling problems from specific domains in a more realistic manner and solving them efficiently. The multiobjective search model is one such framework for modeling and solving search problems involving multiple, conflicting and non-commensurate optimization criteria.
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 3. Multiobjective State Space Search
Abstract
The multiobjective search model appears to be an interesting general framework for solving multicriteria optimization problems. However several practical aspects need to be considered before actually applying the scheme to real world problems. For the conventional search model, extensive work has been done on developing feasible and effective search strategies. Algorithmic improvements on the basic search strategies such as A* have also been suggested. The effect of different types of heuristics have been investigated to cover a broader spectrum of search problems.
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 4. Applications of Multiobjective Search
Abstract
This chapter addresses applications of multiobjective search. The approach is two-fold. The first is to demonstrate the modeling and solving of practical multi-criteria optimization problems using the multiobjective search scheme. The other is to compare the relative efficiency of the multiobjective search techniques developed in chapter 3 (such as MOA** and MOMA*0) with straight-forward generalizations of standard single objective search strategies (such as A* and Depth-first Branch and Bound).
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 5. Multiobjective Problem Reduction Search
Abstract
Problem reduction search is a popular scheme for solving problems that can be hierarchically broken down to a conjunction or disjunction of subproblems [69]. Since such problems can be conveniently represented by AND/OR graphs, problem reduction search is often considered to be synonymous with the problem of searching AND/OR graphs. The algorithm AO* is the most well studied strategy for this representation [11, 41, 62, 63, 68, 69].
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 6. Multiobjective Game Tree Search
Abstract
The study of game playing situations and the task of developing intelligent game playing strategies has been one of the most popular topics in artificial intelligence. Besides popular games such as chess, checkers and GO, game playing situations also arise in many problems of practical significance where an adversary is present [75].
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Chapter 7. Conclusion
Abstract
In this book we have considered three different representation schemes within the multiobjective search framework, namely the state space, problem reduction and game tree representations. We have studied each representation individually and have analyzed the possibility of extending standard search techniques to this framework with suitable modifications. Interestingly our attempts in this direction have led to entirely different situations in the three representation schemes.
Pallab Dasgupta, P. P. Chakrabarti, S. C. DeSarkar
Backmatter
Metadaten
Titel
Multiobjective Heuristic Search
verfasst von
Pallab Dasgupta
P. P. Chakrabarti
S. C. DeSarkar
Copyright-Jahr
1999
Verlag
Vieweg+Teubner Verlag
Electronic ISBN
978-3-322-86853-4
Print ISBN
978-3-528-05708-4
DOI
https://doi.org/10.1007/978-3-322-86853-4