To check these models, why don't they put old data, say 1950s and predict 1980s? If it can't predict known data, it isn't trustworthy.
To check these models, why don't they put old data, say 1950s and predict 1980s? If it can't predict known data, it isn't trustworthy.