The Objectives and Priorities for the Azorean Dairy Farmers’ Decisions - The Agricultural Economics of the 21st Century - page 137

MULTI-CRITERIA DECISION METHODS

Multi-Attribute Decision
Techniques

Multiobjective Decision Techniques

Multigoal
Programming
(MP)

Goal Programming (GP) Multiobjective Programming (MOP)

Simplex
Method
with
Multiple
Objetives

Single
Criterion
of
Synthesis
Approach

Interactive
Local
Judgement
Approach

Lexicographical
Goal
Programming

Weighted
Goal
Programming

Weighted
Coefficients
Method

Subordination
Synthesis
Approach

Constraints
method

NISE

Ideal Solution Election

Compromise Programming
(CP)

Interactive Techniques

Method of
Zionts and
Wallenious

Discrete
Approximation

Continuous
Approximation

STEM Others

Fig. 10.1 Multi-Criteria decision methods (Cohon 1978 ; Zeleny 1982 ; Romero and Rehman
1989 ; Romero 1993 ; Poeta 1994 ; and Silva 2001 )

determine a set of efficient solutions, in a first step, and ideal solutions in a second
step, from the first.
If the decision center has to make a decision in a context of multiple goals, then
he should apply GP. This optimization program is developed by minimizing the
deviations between the actually achieved goals and the aspiration levels set
previously.
This minimization process can be achieved by two alternative ways: Lexico-
graphical Goal Programming and Weighted Goal Programming. The first admits
that decision maker is able to define all goals and establish priorities among them.
The second doesn't consider priorities and the goals are simultaneously embraced
on an objective function by minimizing the sum of all deviations between the goals
and aspiration levels. The deviations are subsequently weighted according to the
importance that the decision center assigns to each goal.
Constraints Method; Weighting Coefficients Method; NISE Method (Non Infe-
rior Set Estimation Method), and Simplex Method with Multiple Objectives are the
Multiobjective Programming methods that allow to achieve the efficient solutions.
The first method (Constraints Method) optimizes one of the objectives and the
remaining is incorporated in the constraints set. The Weighting Coefficients
Method is the combination of all objectives into a single function (aggregated
and weighted), associating a weight or weighting coefficient to each one. The
NISE Method is a variant of the Weighting Coefficients Method based on the
attribution of weights for each objective according to the slope of the straight
lines that connect the extreme efficient points. The method of the simplex algorithm
with multiple objectives generates the efficient set through a “jump” from an