Un article qui peut aider à percevoir les idées reçues.
PatrickSince these limit theorems are sometimes held to be the most important and sophisticated fruits of probability theory, we note that they depend crucially on the assumption of independence of different trials. The slightest positive correction between trials i an j, if it persist for arbitrarily large |i - j|, will render these theorems qualitatively incorrect.
-----