Paper Info
Title
Multicriterially Best Explanations
Abstract
Inference to the best explanation, IBE, (or abduction) requires finding the best explanatory hypothesis, from a set of rival hypotheses, to explain a collection of data. The notion of best, however, is multicriterial and the available rival hypotheses might be variously good according to different criteria. Thus, one can view the abduction problem as that of choosing the best hypothesis from among a set of multicriterially evaluated hypotheses - i` as a multiple criteria decision making problem. In the absence of a single hypothesis that is the best along all dimensions of goodness, the MCDM problem becomes especially hard. The Seeker-Filter-Viewer architecture provides an effective and natural way to use computer power to assist humans to solve certain classes of MCDM problems. In this paper, we apply an MCDM perspective to the abductive problem of red-cell antibody identification and present the results obtained by using the S-F-V architecture.
Year
DOI
Venue
2001
10.1007/3-540-45650-3_14
Discovery Science
Keywords
Field
DocType
best hypothesis,abductive problem,single hypothesis,available rival hypothesis,multicriterially best explanations,abduction problem,rival hypothesis,mcdm problem,best explanation,mcdm perspective,explanatory hypothesis
Antibody identification,Data mining,Architecture,Multiple criteria,Multiple-criteria decision analysis,Inference,Computer science,Decision support system,Multicriteria analysis,Artificial intelligence,Machine learning
Conference
ISBN
Citations 
PageRank 
3-540-42956-5
1
0.41
References 
Authors
2
2
Name
Order
Citations
PageRank
Naresh Iyer1234.40
John R. Josephson21003119.16