You are on page 1of 8


Ahmed Farouk Abdul Moneim
Department of Industrial and Management Engineering, Arab Academy of Science and Technology and Maritime
Transport, Aby Qeer,
Alexandria, Egypt

ABSTRACT intricate if the comparison is made on the

basis of multiple criteria.
A new method is presented to derive a priority 2) Capture and assessment of possible
vector W =( w1, w2, ….wN )T defining the ranking of inconsistencies in comparison judgments of
n competing alternatives or criteria from fuzzy more than two factors or alternatives for the
pairwise comparison judgments matrices. The purpose of discarding heavily inconsistent
pairwise comparisons are accepted as linguistic judgments.
evaluations expressing the relative importance of 3) The uncertainty, imprecision and vagueness
pairs of alternatives (i,j). These linguistic of human comparison judgments.
evaluations are quantified in the form of
trapezoidal fuzzy numbers expressed as quadruples Over the last thirty years, numerous valuable
(aij, bij, cij, dij ) in order to model the vagueness and contributions to the study and analysis of these
imprecision in linguistic evaluations. The problem problems were elaborated. These endeavors started
of finding components of priority vector W, given with the development of Analytical Hierarchy
n(n - 1)/2 quadruples (aij, bij, cij, dij ) is formulated Process (AHP) by T.L. Saaty (1977) as a
in the form of an optimization model with a newly mathematical model built to derive priority vectors
proposed objective function. The problem is solved which arrange competing alternatives, factors and/or
by means of a genetic algorithm. The complications criteria from pairwise comparison judgments. The
and inappropriateness in measuring inconsistencies process of deriving an analytical hierarchy starts,
of fuzzy pairwise comparisons, as presented in usually, with forming a pairwise comparison matrix
existing literatures, are treated and resolved in the with elements aij ( i < j ) carrying values of relative
present work by introducing ratios of inconsistency importance of alternative (factor, criterion) i as
index to index of inconsistency of random fuzzy compared to alternative (factor, criterion) j. Since
pairwise comparisons. The proposed method is elements aji = 1 /aij , number of elements necessary
illustrated by numerical examples and compared to and sufficient to form a pairwise comparison matrix
some of the existing methods in literatures. of size n is n(n-1) / 2 . AHP uses a 9 point scale of
importance in finding aij. After building a reciprocal
KEYWORDS matrix, AHP proceeds further to show that priority
Priority Vector, Trapezoidal fuzzy numbers, Genetic vector W= ( w1 , w2, …….. wn )T is the principal Eigen
Algorithm, Random fuzzy pairwise comparisons, vector of the reciprocal matrix. The corresponding
Inconsistency ratios. Eigen value is shown to be equal to n in case of
perfectly consistent pairwise comparison judgments.
1. INTRODUCTION For nearly consistent and inconsistent judgments,
Eigen values are larger than n. Other methods of
Selecting the optimum alternative in multi-criteria deriving priorities are considered in several
decision problems is one of the most crucial literatures. Direct Least Square (DLS) method (Chu
challenges facing decision makers in engineering and et al 1979)and Logarithmic Least Square method
management in different industries and businesses. (Crawford et al 1985) formulated the problem of
These challenges are: deriving priorities in the form of nonlinear programs.
1) The natural limitations of human capability Mikhailov (2003) presented a survey and
to compare or to judge among more than two comparative study of the different methods of
factors or alternatives. It becomes more solution of the problem of deriving priorities from
single-valued or crisp evaluations of pairwise The significant error in this approach is that
comparisons. DLS and LLS methods have the components of a priority vector may have zero values
drawbacks that they have multiple solutions and lack which leads to infinite relative importance. This is in
explicit measures of inconsistency of pairwise a complete contradiction with assumed finite scale of
comparison judgments similar to that clearly adopted relative importance upon which the problem is
in the AHP. The possibility of having inconsistent formulated and solved. This error is clearly noticed
judgments is a pivotal issue calling for serious in the work of Y.C. Erensal et al (2006) in which
concerns since the validity and credibility of some of priority vectors have components with zero
prioritization is, to a great extent, dependant on values. Chang approach in no way permits
consistency of decision maker’s judgments. evaluations of inconsistencies of pairwise judgments,
that is why in Erensal et al (2006), in spite of the
Decision makers, usually, express their relative
obvious inconsistency of pairwise judgment, as will
evaluations linguistically rather than in exact
be shown later in section 4 of the present work, the
numbers given by the 9 point scale in the standard
work did not detect this high inconsistency and
AHP. Linguistic evaluations such as: extremely
proceeded to find a hierarchy and reached to a
important, moderately important and more or less of
conclusion which is naturally questionable.
the same importance, are characterized by inherent
uncertainty, imprecision and vagueness. T. L. Saaty To overcome such complications, Mikhailov
and L. Vargas (1984), treated the problem of (2003) proposed two approaches: 1) Fuzzy
uncertainty and its effect on the stability of rank Preference Programming (FPP) and 2) Modified
order of competing alternatives. They considered the Fuzzy Preference Programming (MFPP). In the first
pairwise comparison estimations and the resulting approach (FPP), the defuzzification technique known
priorities as random variables with given probability as α – cuts is used. Crisp priorities rather than fuzzy
distributions. However, these quantities are priorities are derived from interval judgments
extremely subjective, differ from person to person corresponding to different α – cut levels. Priorities of
and therefore, they cannot be considered random and the same alternative (factor or criterion) at different α
treated statistically by collecting data and deriving – cut levels are then aggregated to obtain resultant
probability distributions describing the behavior of priority. In the second approach (MFPP), a nonlinear
their populations. Fuzzy numbers are considered as optimization model has been formulated and thereby
the most appropriate model to express uncertainty, avoiding the need for using α – cuts which requires a
imprecision and vagueness of decision makers great deal of computations. The MFPP model
judgments. First approaches to solving the problem formulates the problem in a nonlinear program given
of fuzzy prioritization are given in Van Laarhoven et as follows:
al (1983), J. Buckley (1985), C. Boender et al (1989)
Maximize λ
and others. These approaches followed similar
procedures as adopted in the standard Eigen Vector Subject to :
method. However, performing multiple arithmetic
operations such as addition, multiplication and (m ij − l ij )λ w j − w i + l ij w j ≤ 0
division on fuzzy numbers result in fuzzy priorities
with wide spreads due to propagation of fuzziness. (u ij − m ij )λ w j + w i − u ij w j ≤ 0
Obtained fuzzy priorities have almost no practical n
meaning and sometimes they are irrational
(Mikhailov 2003). To the category of researches that
∑i =1
w i = 1 ( i = 1, 2 ,.. n − 1 ), ( j = 2 , 3 ,.. n )
tried to follow similar procedures as in standard AHP
belongs the work of D.Y. Chang (1996). Chang i < j
determined crisp priority vector by performing fuzzy Where,
ordering and evaluating the truth value of the λ is a variable accounting for the degree of
assertion that fuzzy number I is greater than fuzzy
membership of a ratios wi / wj ,
number J as follows:
Truth( I > = J ) = sup min (µ I (x), µ J (y) ) lij , mij , u ij are triangular fuzzy numbers
expressing the pairwise comparison judgment
between alternative i and alternative j,
where, µ I (x), µ J (y) are membership functions
W = (w1 , w2 ,.....wn ) 2) Solving the reformulated problem by use of
is a priority vector,
one of the most powerful search techniques-
n is the number of alternatives (factors or Genetic Algorithms.
criteria) to be ranked. Reformulating the problem enables the writer to
(1) restore back the concept of inconsistency ratio (ICR)
with reference to inconsistencies of random fuzzy
The optimum solution of the above nonlinear
pairwise comparisons. Values of the random index
program is a vector W* , represents the optimum (RI) are computed for different sizes of pairwise
priority vector that leads to a maximum possible judgmental matrices for the possibility of evaluations
degree of membership. As indicated by L. Mikhailov of (ICR).
(2003), λ is the degree of satisfaction and is a
natural indicator of the inconsistency of decision 2. PROBLEM STATEMENT AND
makers judgments. But however, since λ may accept MATHEMATICAL FORMULATION
negative values as well as positive values it becomes
inappropriate to act as a natural indicator of the Given n(n-1)/2 pair wise comparison judgments
inconsistency and to measure the degree of EVij of the relative importance of n alternatives,
* factors or criteria named afterwards as factors. The
satisfaction. In order to render λ positive for
pairwise comparison judgments are linguistic. It is
inconsistent ratios as well, Mikhailov proposed required , at first, to prioritize these n factors and
to introduce tolerance parameters thus adding to determine priority vector W(w1 ,w2 ,..,wn )T and
the difficulty of the problem. The rationality and then evaluate the relative inconsistency (ICR)of
tangibility of inconsistency ratio (ICR) as adopted by judgments of each decision maker (DM).
T.L. Saaty in the standard AHP disappear when
replaced by the indicator λ. The absence of a 2.1. Modeling of Linguistic Judgments
reference to measure relative to which how severe
inconsistency is in a pairwise comparisons – as the Linguistic evaluations are usually characterized by
inconsistency of random comparisons in the standard their vagueness, imprecision and uncertainty. They
AHP – adds to the inappropriateness of λ. The cannot be expressed by crisp numbers as already
formulation in (1) is a nonlinear program which adopted in standard AHP. Random numbers are also
not the proper model because of the subjectivity of
can be solved numerically by commercial
judgments and the absence of statistical data
software such as “Lingo” with declaration of λ necessary to derive probability distributions
as a free variable, but however this formulation describing the uncertainty. Fuzzy numbers are the
is not amenable to direct application of most appropriate model to quantify linguistic
nontraditional optimization techniques such as judgments. Quantification of linguistic judgments is
Genetic Algorithms since the objective (fitness) necessary for proceeding further with the solution of
function is an implicit function of the decision the stated problem and finding the weights wi of the
variables. Recently, these nontraditional n competing factors. Fuzzy numbers are normal
optimization techniques such as Genetic convex fuzzy sets first introduced by L. Zadeh
Algorithms are widely used because of their (1973). Membership functions are used to express the
simplicity, ease of their implementations, degree of belonging of each element to the fuzzy set.
capability to deal with nonlinearities and their Membership functions of fuzzy numbers may be
proven validity. taken as triangular or trapezoidal. In the present
work, trapezoidal membership functions are
The present work is mainly concerned with the considered since triangular are special cases of the
overcoming of the above mentioned complications trapezoidal. Usually, trapezoidal fuzzy numbers are
and shortcomings in deriving priority vectors from given in the form of quadruples EVij (aij, bij, cij, dij ).
fuzzy pairwise comparison judgments. This, in EVij is evaluating the importance of factor i relative
principle, will be accomplished by: to factor j as judged by a decision maker.
1) Reformulating the problem by proposing an
explicit objective function and using
trapezoidal fuzzy numbers.
G (w1 , w2 ,....wn ) = Min (µ12 , ..., µij ,......, µ(n−1)n )
i< j
(i = 1,2,…, n-1), ( j = 2,3,…n).

wi / wi − aij
µij = δ(w / w ) < b + δb < (w / w ) < c +
aij bij cij dij bij − aij i j ij ij i j ij

Figure 1 A membership Function of a Trapezoidal

Fuzzy number
A scale quantifying linguistic judgments into dij − wi / wi
trapezoidal Fuzzy numbers is given in the following + δ( w / w ) > c
dij − cij i j ij

Linguistic judgment Fuzzy number

when i compared with j aij bij cij dij δA are indicator functions given as follows:
Extremely Important 7 8 9 10
Moderately Important 5 6 7 8 δA = 1 when event A occurs
Important 3 4 5 6
Slightly Important
Nearly of Equal Importance
1 /2
δA = 0 otherwise
Table 1 Scale translating linguistic judgments into Conditions aij < bij and cij < dij are necessary
fuzzy numbers
to keep fitness function G( w1, w2,…, wn ) finite.
Since EVji = 1 / EVij, the matrix of pairwise
comparisons is sufficiently defined by n(n-1)/2 Triangular memberships could be accepted as a
elements which is the number of the elements in the special case where, bij = cij . In this case,
upper triangle of the matrix for which ( i < j ), (i =
1,2,…, n-1), ( j = 2,3,…n). δb ij < ( wi / w j ) < cij =0 . Fitness function G( w1,
w2,…, wn ) as given in (2) is obviously a nonlinear
2.2. Mathematical Formulation function in decision variables wi (i = 1,2,…. n). The
Decision Variables: wi - weight of importance of mathematical formulation of the stated problem may
factor i (i = 1,2,…. n), ( 0 < wi < 1 ) then be written in the following form:

Fitness Function: Maximize G (w1 , w2 ,.....wn )

Our objective is to find values of weights wi such n
that values of their ratios wi / wj will be maximizing
their membership in corresponding fuzzy sets EVij. Subject to ∑i =1
wi = 1
Maximum contributions to the value of the objective
or the fitness function occur when ratios wi / wj lie in (3)
the interval from bij to cij. On the other hand,
contributions diminish as ratios wi / wj get values less
than bij or greater than cij . Negative contributions to
the objective function or fitness function occur as 3. SOLUTION OF THE FORMULATED
ratios wi / wj take values less than aij or greater than PROBLEM BY A GENETIC
dij. Next function G( w1, w2,…, wn ) is proposed, in ALGORITHM
the present work, to represent the behavior of the
fitness of a solution W(w1 ,w2, .,wn) to the problem. The optimization problem formulated in (3), could be
effectively and efficiently solved by applying one the
most powerful search techniques- Genetic 1) Build a chromosome, by generating n
Algorithms (GA) originally developed by Holland random numbers uniformly distributed
J.H. (1975) and later refined by Goldberg D.E. between 0 and 1 and normalize these n
(1989) and others. GA outperforms Directed Search random numbers to render them summing to
Techniques because of their capability to explore one. These n normalized random numbers
wider spaces of feasible solutions aiming at a global represent one of the feasible solutions.
optimum solution and thus have higher effectiveness. 2) Substitute for the values of the priorities wi
GA also outperforms methods of Random Search (i = 1,2,….n) in (2) by the values of n
since they exploit obtained good solutions to arrive to normalized random number obtained in step
better solutions and thereby they are more efficient. 1 in order to evaluate the fitness of the
3) Repeat steps 1 and 2 until having an initial
W1 W2 W3 Wn-1 Wn
population of 30 chromosomes (feasible
Figure 2 A chromosome with n genes solutions).
representing a solution to the problem 4) Decide probabilities of crossover Pxover and
mutation Pmut to start reproduction.
GA starts with devising a special coding set of 5) Select two chromosomes from the population
decision variables known as chromosomes as with giving higher chances to those
inspired by theories of evolution and genetics in chromosomes with higher fitness.
biology. Each chromosome represents a solution to 6) Generate y - a continuous random number
the problem by coding all the decision variables. The [0,1] uniformly distributed. If y <= Pxover
chromosome representation, as shown in Figure 2, then the next operation will be crossover
consists of n boxes (genes). Each gene carries a value otherwise it will be copying. In case of
for one of the n decision variables wi . An initial having crossover apply steps 7 and 8
population of 30 to 50 chromosomes are randomly otherwise skip them and go directly to 9
generated. The fitness of each feasible chromosome 7) In case of having crossover operation,
(solution in which the n decision variables sum to uniform crossover is applied since it is found
one) is evaluated by (2). Consecutive generations are more suitable for the problem at hand. In
evolved from the initial population by applying GA uniform crossover, each two corresponding
operators: crossover, mutation and copying. In genes in the two selected in step 5 parent
biological systems, the nature mercilessly selects chromosomes will exchange their values
only the fittest for longer lives and further with 50% probability. Thereby two new
reproduction. In a similar manner, chromosomes are child chromosomes are formed.
selected for crossover and copying in accordance 8) Generate z – a continuous random number
with and proportional to their fitness values.
[0,1] uniformly distributed. If z <= Pmut then
Crossover of two parent chromosomes is operated
perform mutation in the first gene in one of
aiming at having better offspring. Crossover,
the parents otherwise skip and take next
generally is operated with a predetermined
gene. Repeat this step for all genes of the two
probability Pxover raging from 90% to 95%. Copying
child chromosomes. Normalization of the set
is a complementary with crossover event having
of newly obtained gene values after
probability of 5% to 10% and is operated with the
crossover and mutation should be restored
purpose of enriching the population with good
back and then evaluate the fitness of each
solutions. Mutation is a random change of gene
new child. Go to step 10.
values in order to widen the exploration front. This
9) Copying, in the present algorithm, is
operator is necessary to avoid premature convergence
performed simply by replacing the two worst
and falling in a local optimum. However, it is
chromosomes having least fitness by the two
recommended to apply mutation with low probability
selected in step 5 chromosomes.
Pmut from 1% to 10% in order to preserve the
10) By completing the above steps a new
exploitation capability of GA and not becoming just
generation is already obtained Go to step 5
random searches. The Genetic Algorithm, adopted in
for further reproduction
the present work and named as Fuzzy Genetic
11) Repeat until convergence is obtained or
Prioritization (FGP), is described in the following
reaching to a given terminating signal.
12) The solution is then reached by taking values existing methods adopted for evaluation of
of the genes of the fittest chromosome in the inconsistency in case of fuzzy judgments.
last generation.
It is easy to show that the fitness function G( w1,
The described algorithm is implemented in Visual
w2,…, wn ) in (2) has the upper bound of unity in case
Basic for Application (VBA) under Excel.
of perfectly consistent judgments for which
bij ≤ ≤ c ij . As ratios wi / wj get values less
FUZZY PAIRWISE COMPARISON than bij or greater than cij , the value of fitness
JUDGMENTS function decreases. This behavior of G( w1, w2,…,
When constructing a judgment matrix for more than wn ) motivated the writer to propose the following
two competing factors or alternatives for the purpose expression to evaluate the inconsistency index ICI as
of their ranking, there is a possibility of being follows:
inconsistent in DM pairwise comparisons. For
example comparing factor A with factor B, it is
found that A is more important than B (A > B). ICI = 1 − G (w 1 , w 2 ,....... w n ) ≥ 0 (4)
Similarly comparing B with C, it is found that (B >
C). To be consistent A should be rather more As in the standard AHP, inconsistency measure
important than C. But for any reason DM may enter a becomes more tangible if the index ICI is related to a
wrong entry ( A < C ), the comparison judgment random index RI. Next, values of RI for different
becomes severely inconsistent. If the entry is correct matrix sizes (n = 3 to 9 ) have been evaluated by
( A > C ) but without appropriate evaluation of the solving the problem formulated in (3) with random
relative importance, inconsistent comparisons are fuzzy judgments.
also obtained but to a lesser degree. Inconsistency The inconsistency ratio is given as:
increases as the number of factors to be compared
and ranked increases. T.L. Saaty proposed a method ICI
ICR = (5)
to evaluate inconsistency in judgments and showed RI
that the judgment is perfectly consistent if the
maximum Eigen value of the reciprocal matrix is 5. EVALUATION OF THE RANDOM
equal to the matrix size. Inconsistency index ICI = INDEX ( RI )
( Max. Eigen value – Matrix size ) / (Matrix size – 1 ). In Fuzzy numbers (aij, bij, cij, dij ). are generated
order to take a decision to discard one of DM randomly. The GA described in section 3 is run for
matrices or not, ICI should be compared with the seven values of n (n= 3, 4,…, 9 ). For each value of n,
value of ICI of a completely random judgment RI. the GA is run 100 times in order to take an average
Saaty evaluated RI for matrices of different sizes. for the fitness function G( w1, w2,…, wn ) . Having
The ratio of ICI of any DM judgment to RI gives the G, ICI can be calculated by (4). In case of randomly
inconsistency ratio ICR. By the value of ICR, it can generated fuzzy pairwise judgment, RI = ICI. Next
decided to discard a DM judgment or not, depending in table 2, results of these computations are given.
on the number of available DMs and the criticality of
the decision making problem. Acceptable values of
ICR may take values ranging from say 0.01 to 0.. n 3 4 5 6 7 8 9
Since Saaty’s method is applicable only to case of
DM comparison judgments represented by crisp RI 2.62 4.74 6.18 7.2 7.9 9.62 11.28
(single) values, it cannot be applied directly to the
case of fuzzy representations. As already stated in the Table 2 Random Index RI for different sizes n of
fuzzy pairwise judgment matrix
introduction, one of the main concern of the present
work is to generalize the method of finding ICRs in
the standard AHP to cover cases of fuzzy judgments.
This is decided for the purpose of overcoming
complications and inconveniences aroused in the
Present Chang
The first example to be considered is that presented MFPP by
work method
several times by L. Mikhailov and P. Tsvetnov Mikhailov
(FGP) applied by
(2004). The judgment matrix is given as follows: Using
Erensal et
al (2006)
A 1 (2, 3, 4 ) (1, 2, 3 ) w1 0.0473 0.00 0.0447
B (1/4, 1/3, 1 / 2 ) 1 ( 1/3, 1 / 2, 1 ) w2 0.0282 0.12 0.03188
C ( 1/3, 1 / 2, 1 ) (1, 2, 3 ) 1 w3 0.1223 0.31 0.1224
Table 3 w4 0.2602 0.25 0.2578
The solution obtained by applying (FGP), developed w5 0.542 0.32 0.54314
in the present work, and that obtained by (MFPP), w1 / w2 1.676 0.00 1.402
developed by Mikhailov are presented in the w1 / w3 0.387 0.00 0.366
following table w1 / w4 0.1817 0.00 0.1734
Present work w1 /w5 0.0873 0.00 0.0823
(FGP) w2 /w3 0.231 0.387 0.2608
w1 0.53749 0.538 w2 /w4 0.1084 0.48 0.1237
w2 0.17 0.17 w2 /w5 0.052 0.375 0.0587
w3 0.2925 0.292 w3/w4 0.47 1.24 0.474
w1 / w2 3.1616 3.162 w3/w5 0.2256 0.969 0.2251
w1 / w3 1.83758 1.838 w4/w5 0.48 0.7813 0.4746
w2 / w3 0.5812 0.581 G -1.107 -6.8 λ = -1.06
G( w1, w2,…, wn ) 0.83757 λ = 0.838 ICI 2.107 7.8
ICI 0.1624 ICR 0.341 1.262
RI 2.62
Table 6
ICR 0.062
Table 4 standard AHP to be applied to fuzzy pairwise
comparisons. However, Chang derived a crisp
The comparison in tables 4 reveals the coincidence of priority vector by evaluating the truth value of the
the solutions obtained from the two approaches. This, assertion that a fuzzy number I is greater than a fuzzy
evidently proves the validity of the proposed Fuzzy number J as pointed in the introduction. The priority
Genetic Prioritization (FGP) as developed in the vector derived from reciprocal matrix in table 5
present work. The inconsistency ratio ICR, as obtained by Erensal et al (2006) applying Chang’s
proposed in the present work, gives a value of 0.062 method is given in table 6. The same example as
which is fairly low and indicates a tangible measure given in table 5 is solved by the Fuzzy Genetic
of consistency of the pairwise comparison given in Prioritization FGP of the present work and also by
table 3 rather than using λ = 0.838 .The second the Modified Fuzzy Preference Programming MFPP
example is that considered by Erensal et al 2006 and developed by Mikhailov L. (2003). The comparison
solved by the method of D. Chang 1996. As already of the results of the three approaches is given in table
stated in the introduction, method of Chang belongs 6. The comparison reveals the following:
to a category of works that extend the approach of
• The solution of Erensal et al (2006) is not an
Cost Price Quality Flexibility Time optimum solution since better solutions have
Cost 1 1/5,1/3,1 1/7,1/5,1/3 1/5.1/3,1 1/7,1/5,1/3 been obtained by both FGP and MFPP for
Price 1,3,5 1 1/9,1/7,1/5 0.5,1,1.5 1/5,1/3,1
the same reciprocal matrix.
Quality 3,5,7 5,7,9 1 1/7,1/5,1/3 0.5,1,1..5 • The inconsistency ratio ICR as obtained by
Flexibility 1,3,5 0.9,1,1.1 3,5,7 1 1/7,1/5,1/3 the present work for the optimum solution is
3,5,7 1,3,5 0.9,1,1.1 3,5,7 1
0.341. This ratio reveals that pairwise
judgment in table 5 is inconsistent. This
Table 5 result is confirmed by the findings of MFPP
for which λ = - 1.06.
• The coincidence between results of the 8. REFERENCES
approaches FPG of the present work and Boender C., De Graan and Lootsma F. (1989), “Multi
MFPP of Mikhailov is again clearly noticed. criteria decision analysis with fuzzy pairwise
comparisons”, Fuzzy Sets and Systems 29, pp 133 – 143.
Buckley J. (1985), “Fuzzy Hierarchical analysis”, Fuzzy
1) The extension of the standard AHP method, Sets and Systems 17 pp 233 – 247.
as introduced by T.L. Saaty, in which the Brazili L. A. (1997) “Deriving Weights from pairwise
pairwise comparisons are given by crisp comparison matrices, Journal of Operational Research
values, to be applied directly to cases of Society, 48 pp 1226 - 1232
having fuzzy pairwise judgments is not valid
and leads mostly to irrational inconsistent Chang D. (1996) “Application of the extent analysis
method on fuzzy AHP”, European Journal of
solutions. Operational Research” 95, pp 649 - 655
2) The proper handling of problems of deriving Chu A. , Kalaba R. Springarm K. (1979) “A comparison
priorities from fuzzy pairwise judgments is of two methods for determining the weights of belonging
to formulate them as optimization problems. to fuzzy sets, Journal of Optimization Theory and
Fuzzy Preference Programming developed Application, 27 pp 531 – 541.
by Mikhailov is the first trial in this concern.
Crawford G. and Williams C. (1985) “ A note on the
3) Expressing objective functions, in analysis of subjective judgment matrices” Journal of
optimization models, explicitly is an Mathematical Psychology, 29, pp 387 – 405
important prerequisite to rendering the Erensal Y. C., Oncan T. and Demircan M. (2006),
nonlinear programs amenable to applications “Determining Key Capabilities in Technology
of the effective and widespread Management Using Fuzzy Analytic Hierarchy Process:
nontraditional search techniques such as A case study of Turkey” Journal of Information Science”
Genetic Algorithms. 176 pp 2755 – 2770.

4) It has been demonstrated that Genetic Goldberg D.E. (1989), “Genetic Algorithms in search
Algorithms can be easily and transparently optimization and machine learning”, Addison - Wesley,
applied to solve the formulated optimization
models for deriving priorities from fuzzy Holland J.H. (1975), “Adaptation in natural and artificial
pairwise judgments after modification of the systems”, University of Michigan Press, Ann Arbor,
objective (fitness) function and making it Michigan.
explicit function of decision variables wi. The Mikhailov L. (2003) “ Deriving priorities from fuzzy
coincidence of results obtained from the pairwise comparison judgments” Fuzzy Sets and
application of the developed, in the present Systems 134, pp363 – 385.
work, Fuzzy Genetic prioritization FGP and Mikhailov L. and P.Tsvetinov (2004) “Evaluation of
solutions obtained from MFPP by Mikhailov services using a fuzzy analytic hierarchy process”
demonstrates the validity of the proposed Applied Soft Computing, 5 pp 23 – 33.
model FGP.
Saaty T.L. (1977), “A Scaling method for priorities in
5) To have ICR as a tangible measure of hierarchical structures”, Journal of mathematical
inconsistency of pairwise judgments is Psychology 15, pp 234 - 281
crucial for making rational decisions. The Saaty T.L. and Vargas L. G. (1984), “Inconsistency and
measure ICR acts as a screening element by Rank preservation” Journal of mathematical Psychology
means of which seriously inconsistent 28/2
responses to questionnaires could be
Van Laarhoven P. and Pedrycz W. , (1983), “Fuzzy
discarded. This screening is mandatory prior Extensions for Saaty’s priority theory”, Fuzzy Sets and
to getting geometric means in case of having Systems 11, pp 229 – 241.
group of decision makers. The computations
of the random indexes RI for matrices of Zadeh L. A. (1973) “The concept of linguistic variable
and application to approximate reasoning” ERL-M 411,
different sizes with random fuzzy entries, as
Berkeley, California
elaborated in the present work, made it
possible to evaluate ICRs