Advertisement

AND/OR graphs

  • Igor Aleksander
  • Henri Farreny
  • Malik Ghallab
Part of the Robot Technology book series (RT, volume 6)

Abstract

The difference between searching a graph and searching an AND/OR graph resides in the fact that, in the latter case, the PSS rules are no longer bi-univocal applications (transforming a state u into one other state v), but multivocal applications (transforming u into several states v1 . . . vK). For this reason, the expression ‘state decomposition rule’ is introduced. The search problem here is formulated in a similar way to that in the previous chapter. It is based on finding a method that, starting from an initial state uo, for which several decomposition rules are valid, will choose one in particular for application to uo, start at each of the states obtained by this rule, and continue in the same way until all the states that have not been broken down are terminal.

Keywords

Partial Solution Complete Solution Terminal State Search Problem Previous Chapter 
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.

References

  1. Ghallab, M. Optimisation de Processus Decisionnels pour la Robotique Thesis, Univ. Paul Sabatier, Toulouse, 1982.Google Scholar
  2. Kumar, U.; Kanal, L. A general branch and bound formulation for understanding and synthesizing And/Or procedures. Artificial Intelligence 21, 179–198, 1983.CrossRefGoogle Scholar
  3. Martelli, A.; Montanau, U. Additive and/or graphs. Proc. 3rd IJCAI, Stanford, 1973, pp. 1–11.Google Scholar
  4. Vanderburg, G.J. Problem representations and formal properties of heuristic search. Information Sciences 11, 279–307, 1976.CrossRefGoogle Scholar

Copyright information

© Kogan Page Ltd 1986

Authors and Affiliations

  • Igor Aleksander
  • Henri Farreny
  • Malik Ghallab

There are no affiliations available

Personalised recommendations