One more curious lesson was that, in Bayes' original article on his theorem, he discusses how to infer P(heads) from tossing a coin n times. Today, we know that the unif. prior solution is (m+1)/(n+1), where m= # heads. But Bayes didn't. He couldn't solve the integral! Laplace did this later.
February 9, 2025 at 8:17 PM
One more curious lesson was that, in Bayes' original article on his theorem, he discusses how to infer P(heads) from tossing a coin n times. Today, we know that the unif. prior solution is (m+1)/(n+1), where m= # heads. But Bayes didn't. He couldn't solve the integral! Laplace did this later.