Research Aim

In this project, a high speed neural network classifier based on extreme learning machines for multi-label classification problem is developed. Multi-label classification is a superset of traditional binary and multi-class classification problems. The proposed work extends the extreme learning machine technique to adapt to the multi-label problems. As opposed to the single-label problem, both the number of labels the sample belongs to, and each of those target labels are to be identified for multi-label classification resulting in increased complexity.

Multi-label datasets have a unique property called the degree of multi-labelness. The number of labels, the number of samples having multiple labels, the average number of labels corresponding to a particular sample varies among different datasets. Not all datasets are equally multi-label. Two dataset metrics are available in the literature to quantitatively measure the multi-labelness of a dataset. They are Label Cardinality (LC) and Label Density (LD). The properties of two datasets have same LC, but different LD can vary significantly and may result in different behavior of the training algorithm.


Emotion (Multimedia), Yeast (Biology), Scene (Multimedia), Corel5k (Multimedia), Enron (Text), Medical (Text)

Dataset specifications are given below:



The source code for batch learning for multi-label classifier will be uploaded to github soon


More detailed discussions and results are available in the paper:

Multi-Label Classification Method Based on Extreme Learning Machines [PDF]

A High Speed Multi-label Classifier Based on Extreme Learning Machines [PDF]