We propose the use of anytime performance profiles to describe the
computational behaviour of problem solving methods. A performance
profile describes how the quality of the output of an algorithm
gradually increases as a function of the computation time. Such anytime
descriptions of problem solving methods are attractive because they
allow a trade-off to be made between available computation time and
output-quality. It turns out that many problem solving methods found in
the literature have a natural anytime behaviour, which has remained
largely unexploited until now.
In this paper we propose an axiomatic description of performance
profiles. Furthermore, we give a fixed schematic form for these
axiomatic descriptions. Finally, we apply our proposal to a number of
realistic problem-solving methods, namely hierarchical classification
(used in MDX), and parametric design methods from XCON and VT.
@InProceedings{ECAI00,
author = "F. van Harmelen and A. ten Teije",
title = "Describing Problem Solving Methods
using Anytime Performance profiles",
booktitle = "Proceedings of {ECAI}'00",
year = 2000,
pages = "181--186",
address = "Berlin",
month = "August",
keywords = {Approximate Reasoning},
urlPaper = "http://www.cs.vu.nl/~frankh/postscript/ECAI00.pdf"
}