
[full paper] 
Mónica Sanchez, Francesc Prats, Núria Agell, Joseph Aguilar
The study of hybrid connectives linearly compensated H=ëC + (1ë)C* , where C is a tnorm and C* represents the dual connective of C, to define aggregation operators for fuzzy classifications is a key point not only in fuzzy sets theory but also in learning processes. Although these operators are not associative, the fact that they can be decomposed into associative functions, give rise easily to nary aggregation functions by straightforward iteration. Among the most used tnorms there are those of Frank's family, which are simultaneously tnorms and copulas. The purpose of this paper is to give a characterization of the hybrid connective H, via the properties of the connective C. Necessary and sufficient conditions on H that define C as a copula are given. The characterized hybrid connectives H are used to compute global adequacy degree of an object in a class from marginal adequacy degrees in a learning system.
Keywords: Hybrid Connectives, Reasoning under Uncertainty, Machine Learning, Classification Algorithms, Qualitative Reasoning
Citation: Mónica Sanchez, Francesc Prats, Núria Agell, Joseph Aguilar: Evaluating Global Adequacy in Fuzzy Classifications. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.10811082.