where (.) in this case denotes the components of x being arranged in non-decreasing
order such that (x(1) ≤ x(2) ≤ • • • ≤ x(n)) (note that this is opposite to OWA).
Special cases of the Choquet integral include weighted arithmetic means and the
OWA function where the fuzzy measure is additive or symmetric respectively. Submodular fuzzy measures result in Choquet integrals which are concave, the upshot
of which is that increases to lower inputs affect the function more than increases
to higher inputs. Conversely, supermodular fuzzy measures result in convex functions. Choquet integrals are idempotent, homogeneous, shift-invariant and strictly
monotone where A B → v(A) < v(B). Where the fuzzy measure is symmetric, the
function will obviously satisfy the symmetry property.