Suppose I have an attribute "A" with 4 levels for whom we have not set proiri ranking in an ACA exercise(i.e. it is not ordinal attribute)
However during ACA/HB estimation we want U(A_2) > U(A_3) this constraint to be satisfied keeping no restriction over level 1 and level 4
In ACA/HB, Can we use partial constraint on the utility within an Attribute for some of its levels and not all like the way we can do in CBC-HB?.