抄録
Regional ranking can be made by comparing with regions with each other at various indexes.
However these rankings are not so trustworthy, especialy when there are so many regions that have similar indexes. Therefore rankings are used sometimes trustingly and sometimes doultbully.
As a meams of representing the reliablity of the regional rankings, fuzzy membership functions are utilized.
The membership functions around every index figure are as follows.
g means how near two indexes are. And Gi means how many regions are members of i-region, as it were, “potential of short distance membership”.
Then we difine Φ as follows.
Φ means total difference of “potential”
And Φ is the function of ω as follows.
So Φ has maximum at some point of ω. (_??_Φ_??_0)
As Φ is total difference of potential, it's important to select ω which bring maximum value of Φ. Because the more the value is high, the cleaerer the difference between every Gi.
We select such ω and then observe which regions are near with each other and in those near group we find they has no meaning to rank each other.
