Abstract: Owing to the uncertainty of multi-label data and noise data,a novel multi-label robust fuzzy rough classification model was proposed,which is an extension of k-mean robust statistics fuzzy rough classification model that is used to solve the single label classification problem.First,for each unlabeled instance,the membership with respect to each label was obtained by similarity measures.Second,according to the membership,the degree of correlation was defined.Finally,an appropriate threshold was given to demarcate the correlated and uncorrelated labels. The experimental results on three benchmark multi-label datasets and one actual multi-label datasets indicate that the proposed model is superior to ML-kNN and rank-SVM across six popular multi-label evaluation metrics.

Key words: Fuzzy rough set,k-mean robust statistics,Membership,Multi-label learning