Abstract
Conventionally, we use a chi-square test of homogeneity to determine whether the cell probabilities of a multinomial are equal. However, this process of testing hypotheses is based on the assumption of two-valued logic. If we collect questionnaire data using fuzzy logic, i.e., we record the category data with memberships instead of with a 0-1 type, then the conventional test of goodness-of-fit will not work. In this paper, we present a new method, the fuzzy chi-square test, which will enable us to analyze those fuzzy sample data. The new testing process will efficiently solve the problem for which the category data are not integers. Some related properties of the fuzzy multinomial distribution are also described.
| Original language | English |
|---|---|
| Pages (from-to) | 7437-7450 |
| Number of pages | 14 |
| Journal | International Journal of Innovative Computing, Information and Control |
| Volume | 8 |
| Issue number | 10 B |
| Publication status | Published - 2012 Oct |
Keywords
- Chi-square test for goodness-of-fit
- Fuzzy numbers
- Fuzzy set theory
- Membership functions
- Sampling survey
ASJC Scopus subject areas
- Computational Theory and Mathematics
- Information Systems
- Software
- Theoretical Computer Science