As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
Multi-label classification deals with problems where each of the data instances has several labels associated with it. Although many ensemble-based approaches for multi-label classification have been proposed, several of them do not take into account intrinsic characteristics of the data during their design. In this paper we present a cooperative coevolutionary algorithm which considers such specific characteristics to build an ensemble of accurate and diverse multi-label classifiers. The algorithm evolves several subpopulations simultaneously, each using a different subset of the training data. Also, each individual is focused only on a small subset of labels. These two characteristics provide greater diversity of members to generate the ensemble. As it evolves separate members, we also define a procedure to build an ensemble given the individuals. The experimental study comparing the proposed method to the state-of-the-art in multi-label classification using thirteen datasets and five evaluation metrics demonstrated that the developed cooperative coevolutionary algorithm performed consistently and statistically better than the other methods.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.