A comparison of three problem-solving methods

  • Authors:

  • Ranan B. Banerji;George W. Ernst

  • Affiliations:

  • Temple University, Philadelphia, Pennsylvania;Case Western Reserve University, Cleveland, Ohio

  • Venue:
  • IJCAI'77 Proceedings of the 5th international joint conference on Artificial intelligence - Volume 1
  • Year:
  • 1977

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper is a comparison of ABSTRIPS, planning (as defined in Newell and Simon, 1972) and GPS. Each of these methods has parameters chat contain heuristic information which is problem dependent. These parameters are used to guide the methods, search and usually cause them to be incomplete in the sense that they cannot solve some problems that have solutions. We show that the parameters of the methods serve the same function in the following sense: Given the parameters for one method we can formulate the parameters for the other two such that all three can solve the same class of problem; i.e. those which have totally ordered solutions. This result is somewhat surprising because the search spaces of the methods are different. The Implications of this result to the efficiency of search Is discussed at some length.