In this paper, a high-speed online neural network classifier based on extremelearning machines for multi-label classification is proposed. In multi-labelclassification, each of the input data sample belongs to one or more than oneof the target labels. The traditional binary and multi-class classificationwhere each sample belongs to only one target class forms the subset ofmulti-label classification. Multi-label classification problems are far morecomplex than binary and multi-class classification problems, as both the numberof target labels and each of the target labels corresponding to each of theinput samples are to be identified. The proposed work exploits the high-speednature of the extreme learning machines to achieve real-time multi-labelclassification of streaming data. A new threshold-based online sequentiallearning algorithm is proposed for high speed and streaming data classificationof multi-label problems. The proposed method is experimented with six differentdatasets from different application domains such as multimedia, text, andbiology. The hamming loss, accuracy, training time and testing time of theproposed technique is compared with nine different state-of-the-art methods.Experimental studies shows that the proposed technique outperforms the existingmulti-label classifiers in terms of performance and speed.
展开▼
