Classifying multilabeled data points using Mulan