A statistical model is identifiable when distinct parameter values determine distinct probability distributions. In a nonidentifiable parameterization, data can at most estimate combinations of parameters that remain unchanged along each equivalent parameter set.
Articles by others on the same topic
There are currently no matching articles.