8 ms·

I'm inclined to agree with you here. The solutions suggested here seem to depend on heuristics that depend on the type of distribution. A tangential question:

• scapp 4y ago
  First, a point of terminology: uncountably infinite refers to any infinite cardinality other than countably infinite, so if the number of probability distributions is at least the cardinality of the real numbers, then it's already uncountable. As for what uncountable cardinality they form, it's the same as the real numbers [0]. Roughly, a probability distribution is determined by the countable collection of real numbers P(X <= q) with q rational. That means the cardinality is no more than R^Q, which is isomorphic to R. (the cardinality is at least R due to, say, uniform distributions on [0, x] for x real). [0] https://math.stackexchange.com/questions/3698864/why-does-the-number-of-possible-probability-distributions-have-the-cardinality-o/3698870#3698870 https://math.stackexchange.com/questions/3698864/why-does-th...