Characterization of measures based on strict triangular norms (Q1304687): Difference between revisions

Let \({\mathcal B}\) be a \(\sigma\)-algebra of subsets of \(X\) and \({\mathcal T}\) be the tribe of all \({\mathcal B}\)-measurable functions \(A: X\to [0,1]\) (fuzzy subsets of \(X\)). Let \(T\) be a t-norm, i.e., a binary function \(T:[0, 1]^2\to [0,1]\) which is commutative, associative, non-decreasing and satisfies the boundary condition \(T(a,1)= a\) for all \(a\in [0,1]\). \(T\) is called strict if \(T\) is continuous and \(a>0\) and \(b<c\) implies \(T(a, b)< T(a,c)\). The associated t-conorm \(S\) is defined by \(S(a,b)= 1- T(1- a,1- b)\). A \(T\)-measure \(n:{\mathcal T}\to \mathbb{R}\) is a left continuous function satisfying \(m(0)= 0\) and \(m(T(a, b))+ m(S(a, b))= m(a)+ m(b)\). The author obtains the following surprising result: Let \(T\) be a strict t-norm. Then any monotone \(T\)-measure is of the form \(m(A)- \mu(\text{supp }A)+ \int h\circ A d\nu\), where \(\mu\), \(\nu\) are (classical) measures on \({\mathcal B}\) and \(h:[0, 1]\to [0,1]\) is an order automorphism of \([0,1]\) or \(h= 0\).