Abstract
This research study deals with the improvement of multi-label classification by modeling existing dependencies between labels. The main purpose of
the study is to define and develop a classification algorithm for multi-label classification tasks by partitioning the class set into several subsets. According to this
algorithm, first, dependencies among the labels are analyzed and then the whole
set of labels is divided into several mutually exclusive subsets. Finally, a classification algorithm incorporating dependencies among labels within each subset
can be applied. An experimental study shows that the proposed method has high
potential to achieve the defined objectives and improve multi-label classification
performance.
the study is to define and develop a classification algorithm for multi-label classification tasks by partitioning the class set into several subsets. According to this
algorithm, first, dependencies among the labels are analyzed and then the whole
set of labels is divided into several mutually exclusive subsets. Finally, a classification algorithm incorporating dependencies among labels within each subset
can be applied. An experimental study shows that the proposed method has high
potential to achieve the defined objectives and improve multi-label classification
performance.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the 1st international workshop on learning from multi-label data, Bled, Slovenia |
| Editors | Grigorios Tsoumakas, Min-Ling Zhang, Zhi-Hua Zhou |
| Place of Publication | Bled, Slovenia |
| Pages | 117-132 |
| Number of pages | 16 |
| State | Published - Aug 2009 |