Advertisement

Abstract

To my knowledge, the exact problem formulation I adopted for this dissertation — satisfying hidden strong constraints — had not previously been studied. This makes comparisons with other work somewhat problematical, since as suggested in Chapter 1, even small differences in problem formulations can have large impacts on the appropriate choice of problem-solvers. Consequently, this chapter begins by briefly considering the relationships between the adopted problem formulation and one widely-recognized approach, the problem space formulation. It is difficult to make precise comparisons without a common “meta problem formulation” to compare both approaches to, but a tentative conclusion is offered that the two approaches are largely complementary along a spectrum defined by the size of the average branching factor of a search tree. When a problem presents a lot of “inherent sequentially”—a relatively deeper search tree—the problem space approach applies more naturally; when a problem is “wide open”—a relatively bushier search tree—the function optimization approach applies more naturally.

Keywords

Genetic Algorithm Simulated Annealing Receptive Field Input Pattern Learning Rule 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Kluwer Academic Publishers 1987

Authors and Affiliations

  • David H. Ackley
    • 1
  1. 1.Carnegie Mellon UniversityUSA

Personalised recommendations