Admissibility

A very common concept which is applicable to any procedure of the statistical inference. The underlying notion is that the procedure/method is admissible if and only if there does not exist within that particular class of the procedures another one which performs uniformly at least as well as the procedure in question and performs much better than it in at least one case. Here 'uniformly' means for all values of the parameters which decide the probability distribution of the random variables under the investigation.