For example, I want to use some linguistic terms (e.g., young, middle age, and old) for dividing our ages into some fuzzy categories. In order to use fuzzy c-means on discretized data set.
For numerical attribute like age, we can do fuzzy discretization in following steps.
(i) Decide number of fuzzy sets (in your example it is 3 i.e. young, middle-age and old).
(ii) Select membership function for fuzzy set.
(iii) Membership function is to be such that there will be different membership value in different fuzzy set. For ex., age 20 can be in fuzzy set young with membership value 0.8, in fuzzy set middle-age with membership value 0.3 and in fuzzy set old with membership value 0.
(iv) Replace each age value in original data by corresponding linguistic term (name of fuzzy set) with membership value.
The idea of fuzzy discretization is to make overlapping clusters of fuzzy sets.
hello friend
first you need to apply fuzzification and then you apply defuzzification...
see my paper for more details.
Best regards
The following links may be helpful for you,
http://people.na.infn.it/~lizzi/seminari/03napoli.pdf
http://www.wseas.us/e-library/conferences/2006cavtat/papers/525-104.pdf
Hello Omar,
you question is an classical one and therefore, you will find answers in e.g.
George J Klir, Bo Yuan, Fuzzy sets and Fuzzy Logic, Theory and Applications, Prentice Hall of India (1997).
They used exactly what you are looking for as an example. However, there are some general aspects in fuzzyfication of verbs, adverbs etc. For a first guess I attach a section of my lecture regarding this topic. Maybe it helps....
the data types of much practical data are continuous or hybrid. instead of the discretization of continuous data we can fuzzify the data.
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6622533&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6622533
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining