Continuous R-implications generated from representable aggregation functions (Q991413)

Representable aggregation functions are non-associative and conjunctive binary operations on the interval \([0,1]\) based on the functional \(F(x,y)=g^{-1}(\max(0,g(x \land y)-g(N(x \lor y))))\), where \(N\) is a strong negation and \(g:[0,1]\to[0,\infty]\) is a continuous strictly increasing function with \(g(0)=0\). In this article, the authors introduce R-implications derived from this class of functions, i.e., functions based on the functional \(I_F(x,y)=\sup\{t\in [0,1] \;| \;F(x,t) \leq y\}\). A detailed study of such implications is given, where they are proved to be continuous fuzzy implications in the sense of Fodor and Roubens. Several important properties, like neutrality principle, ordering property, contrapositive symmetry and exchange principle, are analyzed. This allows the authors to obtain new solutions of the modus ponens and modus tollens properties with nilpotent t-norms.