In statistics, identifiability is a property which a model must satisfy in order for inference to be possible. We say that the model is identifiable if it is theoretically possible to learn the true value of this model’s underlying parameter after obtaining an infinite number of observations from it.
详细百科解释