AN UNBIASED VIEW OF REGRESSION

It follows from your LLN that if an event of probability p is noticed frequently throughout unbiased experiments, the ratio on the observed frequency of that event to the whole number of repetitions converges to p.Other distributions may well not even be a mix, by way of example, the Cantor distribution has no beneficial probability for almost any