and our prediction by is the most common value of Y for the observed data