I am interested in knowing specific cases where a numerical algorithm acquires the status of a full proof (which makes it automatically become a theorem.) How do you verify rigorously that the designed algorithm is actually the proof you were looking for?

Similar questions and discussions