Resampling in Fuzzy Regression via Jackknife-after-Bootstrap (JB)

In fuzzy regression modeling, when the sample size is small, resampling methods are appropriate and useful for improving model estimation. However, in the commonly used bootstrap method, the standard errors of estimates are also random because of randomness existing in samples. This paper investigates the use of Jackknife-after-Bootstrap (JB) in fuzzy regression modeling to address this problem and produce estimates with smaller mean prediction errors. Performance analysis is carried out through some numerical illustrations and some interactive graphs to illustrate the superiority of the JB method compared to the bootstrap. Moreover, it is demonstrated that using the JB method, we have a significant model, with some sense; however, this is not the case using the bootstrap method.