Analysis of Verbal Decision Analysis Methods Considering a Multicriteria Model to the Diagnosis of Alzheimer’s Disease I. Tamanini 1 , P. R. Pinheiro 1 , and M. C. D. Pinheiro 1 1 Graduate Program in Applied Computer Sciences, University of Fortaleza (UNIFOR), Fortaleza, Ceará, Brazil Abstract— The Verbal Decision Analysis framework in- volves two ordering methods: PACOM and ZAPROS, being each method best suited for a determined context. The PA- COM method is indicated to problems having a small set of alternatives, being possible to solve models presenting many criteria. The ZAPROS method, though, should be applied to problems having a small criteria set, but presenting a great set of alternatives. Our proposal is to analyse both methods considering their advantages and disadvantages. An approach structured on the ZAPROS method but with some modifications to improve the alternatives comparison process will be also considered. A multicriteria model aim- ing at establishing which test from a determined set would be more likely to lead to the diagnosis of the Alzheimer’s disease faster will be applied on the analysis. The model was previously structured and studied, and the results obtained by the application of each method will be exposed and discussed. Keywords: Verbal Decision Analysis, Ordination Methods, Diag- nosis of Alzheimer’s Disease, Neuroimaging 1. Introduction One of the greatest problems faced in organizations is related to the decision making process. The determination of the object which will conduct to the best result is not a trivial process and it involves the analysis of several factors. These problems are complex and the consideration of all relevant aspects to the decision making is practically impossible, due to the human limitations. Besides, the decision making related to management decisions is even more critical, since the choice of an inaccurate alternative may lead to a waste of resources, affecting the company. Multicriteria methodologies help to generate knowledge about the decision context, thus, increasing the confidence of those who make decisions on the results [5]. We have mul- ticriteria methods based either on quantitative or qualitative analysis of the problem and it is challenging to choose the approach that best fits the problem to be solved. Illustrating problems that apply the quantitative methods, we have [3], [4], [17], [14], and, among those that apply qualitative ones, [12], [20], [21] and [22]. This work focuses on the comparison of the applications of the methods PACOM [9], ZAPROS III [11], and of a method mainly structured on the ZAPROS method [23][19]. The methods above mentioned belong to the Verbal Decision Analysis framework [6], and their main characteristic is that the problems’ solving is performed in a more realistic way from the decision maker’s point of view (quantitative methods could lead to loss of information when one tries to assign exact measures to verbal values). These methods were chosen among others because this paper aims at performing a comparative study of qualitative ordination methods. 2. The PACOM Method The PACOM (PAired COMpensation) method [9] belongs to the Verbal Decision Analysis framework. The method aims at supporting the decision making process in scenarios involving a rather small set of alternatives (no more than 10) and on which the latter are difficult to be analyzed without a formal method. In problems where the number of alternatives is greater than 10, the set should be reduced by applying more robust methods to determine a subset containing the potentially better alternatives. First, a pair of alternatives is selected for comparison and the decision maker is asked to rank order the disadvantages of an alternative with respect to another, from the most significant disadvantages to the least one. Then, hypothetical alternatives are structured to the comparison and compen- sation of part of the alternatives’ disadvantages (the hypo- thetical alternative differ from the real alternative in two or three criteria values, assigning to these the values presented in the other real alternative being compared). Later, the compensation process of two alternatives (generally, one real and one hypothetical) starts. The definition of hypothetical alternatives and the compensation process is performed until the comparison between the real alternatives selected can be established based on the decision maker’s preferences informations obtained on the former comparisons. The procedure of preferences elicitation is psychologi- cally valid, and it is given by the ordering of the relative disadvantages between pairs of real alternatives and by the comparison of auxiliary alternatives. It is based on the pairs compensation principle, implying that the disadvantages of an alternative may be counterbalanced by the disadvantages of another one and, then, the analysis of the more important advantagesdisadvantages is performed. If the disadvantages of the an alternative are less important to the decision maker than the disadvantages of another alternative, than