A model of costs and benefits of meta-level computation
It is well known that meta-computation can be used to guide other computations (at the object-level), and thereby reduce the costs of these computations. However, the question arises to what extent the cost of meta-computation offsetts the gains made by object-level savings. In this paper we discuss a set of equations that model this trade-off between object-savings and meta-costs. The model shows that there are a number of important limitations on the usefulness of meta-computation, and we investigate the parameters that determine these limitations.
Unable to display preview. Download preview PDF.
- 2.H. Lowe. Empirical evaluation of meta-level interpreters. Master's thesis, Department of Artificial Intelligence, University of Edinburgh, 1988. (M.Sc. Thesis).Google Scholar
- 3.S. Minton. Quantitative results concerning the utility of explanation-based learning. Artificial Intelligence, 42:363–392, 1990.Google Scholar
- 4.R. A. Okeefe. The Craft of Prolog. MIT Press, Massachussetts, 1990.Google Scholar
- 5.S. Owen. The development of explicit interpreters and transformers to control reasoning about protein topology. Technical Memo HPL-ISC-TM-88-015, Hewlett-Packard Laboratories Bristol Research Centre, Bristol, U.K., 1988.Google Scholar
- 6.J.S. Rosenschein and V. Singh. The utility of meta-level effort. Report HPP-83-20, Stanford Heuristic Programming Project, March 1983.Google Scholar
- 7.D. von Winterfeldt and Ward Edwards. Decision Analysis and Behavioral Research. Cambridge University Press, 1986.Google Scholar