Abstract
The aim of this chapter is to expose a multi-ways fuzzy analysis of variance (Mult-FANOVA) approach when the fuzziness is taken into consideration. This latter is based on a fuzzy regression model. As such, we expose two exploratory approaches: the first one by preserving the fuzzy nature of the calculated sums of squares using fuzzy approximations and afterward by defining a related heuristic decision rule to be able to decide, and the second one using a particular metric all along the calculations, namely the signed distance and the generalized one. For both approaches, the bootstrap technique is used. A procedure of calculating sequential sums of squares is also exposed when the ordering effect intervenes. For sake of illustration, we give two detailed empirical applications of our approach, where both are composed by multiple factors. By these applications, the objective is to present the Mult-FANOVA model where the generalized signed distance is used, and to compare it with the analysis using the same setups but in the classical environment. One of our main findings is that in both cases we get the same decision of rejecting or not the hypothesis of equality of the population mean vectors, but the decomposition of the sums is different. For the analysis on the residuals of the constructed model, the normality of their distributions is rejected, but stronger in the classical approach. This result can be seen as a disadvantage of the classical approach, and by consequence strengthen our support to the fuzzy one.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
Despite the complexity induced by the difference arithmetic in the decomposition of \(\tilde {X}_{i_1 i_2 j}\), remark that \(\overline {\tilde {X}}_{i_1 i_2 \bullet }\) is expressed by this form in accordance with the expression of the classical theory.
- 3.
The case with interactions can be similarly conceivable. Models which include interactions can be found in Berkachy (2018).
- 4.
For more information about this test, Tukey (1949) described the detailed procedure.
References
Berkachy, R. (2018). Fuzzy Multivariate Analysis of Variance using the Signed Distance; Application on a Data Set, Nestlé Beverages, Technical Report Nestlé beverages Orbe.
Berkachy, R., & Donzé, L. (2018). Fuzzy one-way ANOVA using the signed distance method to approximate the fuzzy product. In Collectif LFA (Eds.), LFA 2018 - Rencontres francophones sur la Logique Floue et ses Applications Cépaduès (pp. 253–264).
Bonferroni, C. E. (1935). Il calcolo delle assicurazioni su gruppi di teste. Tipografia del Senato. https://books.google.ch/books?id=xnQ4cgAACAAJ
Buckley, J. J., & Jowers, L. J. (2008). Fuzzy linear regression I. In Monte Carlo Methods in Fuzzy Optimization (pp. 117–125). Berlin: Springer. ISBN: 978-3-540-76290-4. https://doi.org/10.1007/978-3-540-76290-4_11
Celmins, A. (1987). Least squares model fitting to fuzzy vector data. Fuzzy Sets and Systems, 22(3), 245–269. ISSN: 0165-0114. https://doi.org/10.1016/0165-0114(87)90070-4. http://www.sciencedirect.com/science/article/pii/0165011487900704
Diamond, P. (1988). Fuzzy least squares. Information Sciences, 46(3), 141–157. ISSN: 0020-0255. https://doi.org/10.1016/0020-0255(88)90047-3. http://www.sciencedirect.com/science/article/pii/0020025588900473
Gil, M. A., Montenegro, M., González-Rodríguez, G., Colubi, A., & Casals, M. R. (2006). Bootstrap approach to the multi-sample test of means with imprecise data. Computational Statistics & Data Analysis, 51(1), 148–162. ISSN: 0167-9473. https://doi.org/10.1016/j.csda.2006.04.018. http://www.sciencedirect.com/science/article/pii/S0167947306001174
Gonzalez-Rodriguez, G., Colubi, A., & Gil, M. A. (2012). Fuzzy data treated as functional data: A one-way ANOVA test approach. Computational Statistics & Data Analysis, 56(4), 943–955.
Gonzalez-Rodriguez, G., Colubi, A., Gil, M. A., & D’Urso, P. (2006). An asymptotic two dependent samples test of equality of means of fuzzy random variables. In: Proceedings of the 17th IASC-ERS.
Guo, P., & Tanaka, H. (2006). Dual models for possibilistic regression analysis. Computational Statistics & Data Analysis, 51(1), 253–266. ISSN: 0167-9473. https://doi.org/10.1016/j.csda.2006.04.005. http://www.sciencedirect.com/science/article/pii/S0167947306000910
Hotelling, H. (1951). A generalized T test and measure of multivariate dispersion. In: Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability (Univ of Calif Press, 1951) (pp. 23–41).
Jiryaei, A., Parchami, A., & Mashinchi, M. (2013). One-way Anova and least squares method based on fuzzy random variables. TJFS: Turkish Journal of Fuzzy Systems, 4(1), 18–33.
Kang, M.-K., & Han, S.-I. (2004). Multivariate analysis of variance for fuzzy data. International Journal of Fuzzy Logic and Intelligent Systems, 4(1), 97–100.
Lawley, D. N. (1938). A generalization of Fisher’s Z test. Biometrika, 30(1–2), 180–187. ISSN: 0006-3444. https://doi.org/10.1093/biomet/30.1-2.180. http://oup.prod.sis.lan/biomet/article-pdf/3/1-2/180/422459/30-1-2-180.pdf
Montenegro, M., Gonzalez-Rodriguez, G., Gil, M. A., Colubi, A., & Casals, M. R. (2004). Introduction to ANOVA with fuzzy random variables. In Soft methodology and random information systems (pp. 487–494). Berlin: Springer.
Nather, W. (2006). Regression with fuzzy random data. Computational Statistics & Data Analysis, 51(1), 235–252. ISSN: 0167-9473. https://doi.org/10.1016/j.csda.2006.02.021
Parchami, A., Nourbakhsh, M., & Mashinchi, M. (2017). Analysis of variance in uncertain environments. Complex & Intelligent Systems, 3(3), 189–196
Pillai, K. C. S. (1965). On the distribution of the largest characteristic root of a matrix in multivariate analysis. Biometrika, 52, 405–414. https://doi.org/10.1093/biomet/52.34.405
Ramos-Guajardo, A. B., & Lubiano, M. A. (2012). K-sample tests for equality of variances of random fuzzy sets. Computational Statistics & Data Analysis, 56(4), 956–966. ISSN: 0167-9473. https://doi.org/10.1016/j.csda2010.11025. http://www.sciencedirect.com/science/article/pii/S0167947310004536
Tanaka, H., Uejima, S., & Asai, K. (1982). Linear regression analysis with fuzzy model. IEEE Transactions on Systems Man, and Cybernetics, 12(6), 903–907. ISSN: 0018-9472. https://doi.org/10.1109/TSMC.1982.4308925
Tukey, J. W. (1949). Comparing individual means in the analysis of variance. Biometrics, 5(2), 99–114. ISSN: 0006-341X, 1541-0420. http://www.jstor.org/stable/3001913.
Wilks, S. S. (1932). Certain generalizations in the analysis of variance. Biometrika, 24, 471–494. https://doi.org/10.1093/biomet/24.3-4.471
Wilks, S. S. (1946). Sample criteria for testing equality of means equality of variances and equality of covariances in a normal multivariate distribution. The Annals of Mathematical Statistics, 17(3), 257–281. ISSN: 00034851. http://www.jstor.org/stable/2236125
Wu, H.-C. (2007). Analysis of variance for fuzzy data. International Journal of Systems Science, 38(3), 235–246.
Author information
Authors and Affiliations
Appendix
Appendix
This Appendix is devoted to the description of the Nestlé data set as follows:
1.1 D Description of the Nestlé data base
This data base is collected by the beverage section of the company Nestlé. An expert is asked to rate different aspects of a given confidential beverage. The collected data set is composed by 160 observations resulting from 32 trials of an experiment on the beverage. The full data base comprised of 13 variables among them 11 factors having 2 levels each given by “Up” and “Down,” and the variables FLOW and FILLING. A description of the variables constituting this data set is given in Table 8.12.
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Berkachy, R. (2021). Fuzzy Analysis of Variance. In: The Signed Distance Measure in Fuzzy Statistical Analysis. Fuzzy Management Methods. Springer, Cham. https://doi.org/10.1007/978-3-030-76916-1_8
Download citation
DOI: https://doi.org/10.1007/978-3-030-76916-1_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-76915-4
Online ISBN: 978-3-030-76916-1
eBook Packages: Business and ManagementBusiness and Management (R0)