An approach to test the crisp statistical hypotheses upon the fuzzy test statistic is developed. The proposed approach is investigated for two cases: the case without nuisance parameter, and the case with a nuisance parameter. For rejecting or accepting the null hypothesis of interest, an evidential point of view is proposed, based on a concept called degree of consistency of observations with the hypothesis. The approach is applied to test the parameters of the normal, Bernoulli, Poisson, and exponential distributions, as well as to test the difference of means and the ratio of variances of two normal distributions.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。