Information Technology Reference
In-Depth Information
Elements of a matrix of pairwise similarity and pairwise consistency of expert
results are presented below.
Elements of a matrix of pairwise similarity
1
0,857
0,823
0,857
1
0,863
0,823 0,863
1
Elements of a matrix of pairwise consistency
1
0,684
0.635
0,684
1
0,685
0,635
0,685
1
The matrix of pairwise similarity allows to conclude that results of the 2nd and
3rd experts are of the greatest similarity. It means that they not only refer the most
representative part of the considered group to the same terms, but also
representatives within the terms mainly coincide. The matrix of pairwise
consistency allows to conclude that the greatest consistency is for results of 2nd
and 3rd experts. These results are similar not only within the limits of a great bulk
of plants, but also are uniform for all representatives of the group considered.
Matrixes of pairwise similarity and pairwise consistency are matrixes of fuzzy
relations of similarity
R
and
R
over set of the formalized expert evaluations,
accordingly. To construct fuzzy relations of similarity, transitive closures
R
and
R
are to be defined. Outcomes of the fuzzy cluster analysis of the
formalized expert results under similarity relation
R
and
R
—
2
R
are presented below.
R
and
2
1
Fuzzy clusterization of expert evaluations
α
-level
Cluster
R
Similarity
0,857
0863
1
{1,2,3}
{1},{2,3}
{1},{2},{3}
R
Similarity
2
0,684
0,685
1
{1,2,3}
{1},{2,3}
{1},{2},{3}
The obtained results allow concluding that the plants state evaluation results
obtained by 2nd and 3rd experts are of the most similarity.
