Axiomatizations of quasi-Lovász extensions of pseudo-Boolean functions

Abstract

We introduce the concept of quasi-Lovász extension as being a mapping f : I n → R defined on a nonempty real interval I containing the origin and which can be factorized as f (x 1 , . . . , xn) = L(ϕ(x 1 ), . . . , ϕ(xn)), where L is the Lovász extension of a pseudo-Boolean function ψ : {0, 1} n → R (i.e., the function L : R n → R whose restriction to each simplex of the standard triangulation of [0, 1] n is the unique affine function which agrees with ψ at the vertices of this simplex) and ϕ : I → R is a nondecreasing function vanishing at the origin. These functions appear naturally within the scope of decision making under uncertainty since they subsume overall preference functionals associated with discrete Choquet integrals whose variables are transformed by a given utility function. To axiomatize the class of quasi-Lovász extensions, we propose generalizations of properties used to characterize Lovász extensions, including a comonotonic version of modularity and a natural relaxation of homogeneity. A variant of the latter property enables us to axiomatize also the class of symmetric quasi-Lovász extensions, which are compositions of symmetric Lovász extensions with 1-place nondecreasing odd functions.Mathematics Subject Classification (2010). Primary 39B22, 39B72; Secondary 26B35.

Quasi-Lovász Extensions on Bounded Chains

“…The following two results are variants of Lemma 2 and Proposition 3 in [4]. and their proofs follow in complete analogy.…”

“…A generalization of the notion of Lovász extension of a pseudo-Boolean function was introduced in [4] and called "quasi-Lovász extension". In this paper we extend this notion to the following concept.…”

Fuzzy Measures and Integrals: Recent Developments

“…In the discrete case, a characterization using comonotonic modularity was obtained by Couceiro and Marichal [46,47]. The Sugeno integral was characterized in the discrete case by de Campos and Bolaños [44].…”

Construction of a Choquet integral and the value functions without any commensurateness assumption in multi-criteria decision making

“…Situation (ii) in Theorem 1 implies that criteria i and k are independent (see (9)). The condition that…”

“…In this setting, there is no commensurateness issue since all attributes and thus all value functions are identical. In [9], the Choquet integral is axiomatized as an aggregation function in the context of decision under uncertainty. In [10], the neutral element of a bipolar scale is determined under the knowledge that the decision maker satisfies a special decision strategy called conditional relative importance between criteria.…”