Unfortunately we have realized that there are a few incorrect statements in the paper "Decision Making by Credal Nets", IHMSC'11.
We apologize for that, and soon we will make a clear explanation of the issues we have found. In summary, there is a flaw in the
translation of the E-admissibility on page 3 (a max should be a min). Moreover, the paper asserts the hardness of the decision criteria,
which is correct, although the paper itself does not prove the result in a strictly formal way.