Applications of Multi-Label Classification
International Journal of Innovative Technology and Exploring Engineering · 2020
The absence of labels and the bad quality of data is a prevailing challenge in numerous data mining and machine learning problems. The performance of a model is limited by available data samples with few labels for training. These problems are ultra-critical in multi-label classification, which usually needs clean data. Multi-label classification is a challenging research problem that emerges in several applications such as multi-object recognition, text categorization, music categorization and image classification. This paper presents a literature review on multi-label classification, various evaluation metrics used for analyzing performance and research hchallenges.