A likelihood ratio procedure is introduced in order to test statistical hypotheses when the available data are imprecise (fuzzy) rather than crisp. First, we provide a suitable construction for dealing with fuzzy data based on the concepts of fuzzy sample space and fuzzy-valued random sample. Then, we establish the likelihood ratio procedure using the probability distribution associated with these concepts. Some numerical examples are given for testing the parameters of binomial and normal distributions that will provide an illustration of the procedure.
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。