Abstract: Case-based reasoning is a problem solving method that uses
stored solutions to problems to aid in solving similar new
problems. One of the difficulties of case-based reasoning
is identifying cases that are relevant to a problem. If the
problem is defined on a geometric domain --- for instance,
planning a route using a city map --- it becomes possible to
take advantage of the geometry to simplify the task of finding
appropriate cases. We propose a methodology for determining
a set of cases which ... (Update)
...routes. More detailed information regarding case retrieval (including runtime and efficiency results) can be found in a previous paper [8]. Case Representation and Indexing When PRODIGY generates a plan, the detailed derivational trace of the solution produced is stored as...
.... the similarity metric which effectively takes into account the natural geometric and continuous valued characteristics of the map domain [4], and a discussion of the learning opportunities potentially offered by the real execution of the proposed planned routes [5] The main...
Karen Zita Haighand Jonathan Richard Shewchuk. Geometric similarity metrics for case-based reasoning. In Case-Based Reasoning: Working Notes from the AAAI94 Workshop, pages 182--187, Seattle, WA, August 1994. AAAI Press. http://citeseer.nj.nec.com/haigh94geometric.html More
@inproceedings{ haigh94geometric,
author = "Karen Zita Haigh and Jonathan Richard Shewchuk",
title = "Geometric Similarity Metrics for Case-Based Reasoning",
booktitle = "Case-Based Reasoning: Working Notes from the {AAAI}-94 Workshop",
publisher = "(Menlo Park, CA: AAAI Press)",
address = "Seattle, WA",
pages = "182--187",
year = "1994",
url = "citeseer.nj.nec.com/haigh94geometric.html" }