Abstract: Due to the high-dimensional label space, the imbalanced label distribution problem commonly exists in multi-label datasets. The classification performance of multi-label learning can be improved to some extent by taking care of this problem. Improving classification performance through label correlation is one of the most common and effective strategies. Many scholars have done a lot of re- searches, but most of these studies use positive correlation-based strategies to improve performance. In practice, besides positive correlation, negative correlation of labels may also exist. If both positive correlation and negative correlation are considered, the performance of the classifier will undoubtedly be further improved. Therefore, an imbalanced multi-label learning algorithm MLNCE is proposed based on the enhancement of negative correlation. It aims to alleviate the imbalance problem of multi-labels while considering the positive and negative correlation among labels and further improving the classification performance of multi-label classifiers. The algorithm first uses the label density information to transform the label space, then explores the true positive and negative correlation information among labels in the density label space, and adds it to the classifier objective function. Finally, the accelerated gradient descent is used to solve the output weights to obtain the prediction results. The proposed algorithm is compared with six other multi-label learning algorithms on 11 multi-label standard datasets, and the results show that the algorithm can effectively improve the classification accuracy.

Key words: