The formalism of support logic provides a framework for deductiveinference, with mathematically sound and consistent treatment ofuncertainty and evidence which is aggregated through the reasoningprocess. The authors apply support logic programming to patternrecognition. Initially, a pattern classifier is constructed by encodingexpert knowledge of the problem domain into rules of support logic.Fuzzy sets allow the general properties of features to be describedprecisely. Semantic unification provides an alternative to the usualmetric-based similarity criteria. The validity of the approach isestablished by cross-validating the support logic classifier againstmodels from alternative paradigms. The authors then attempt tocircumvent the requirement for a domain expert, and assess the extent towhich data-driven learning processes can be used to automatically derivecomponents of the support logic classifier
展开▼
