Fuzzy Modeling and Genetic Algorithms for Data Mining and Exploration

262

■

Chapter 7 Fuzzy Clustering

are close to the centroid of the cluster. This approach would induce rules

such as the following.

If a is somewhat around C1(a)

and b is somewhat around C1(b)

then c is somewhat near C1(c)

If a is very around C2(a)

and b is very around C2(b)

then c is very near C1(c)

If a is around C1(a)

and b is around C1(b)

then c is near C1(c)

The combined truth of the predicate expressions will cause the rule with

the greatest degree of specificity to contribute more strongly to the out-

come and thus provide a higher degree of evidence for the final outcome.

Other possible approaches to using hedges and contract intensifiers or

dilutors can be employed to produce a wide family of possible rules.

7.9 Review

This chapter introduced and explored ...