public interface IDyadRanker extends IBatchLearner<IDyadRankingInstance,IDyadRankingInstance,DyadRankingDataset>
An abstract representation of a dyad ranker.
"Label ranking is a specific type of preference learning problem, namely the prob- lem of learning a model that maps instances to rankings over a finite set of predefined alternatives. Like in conventional classification, these alternatives are identified by their name or label while not being characterized in terms of any properties or features that could be potentially useful for learning. In this paper, we consider a generalization of the label ranking problem that we call dyad ranking. In dyad ranking, not only the instances but also the alter- natives are represented in terms of attributes."
Schäfer, D., & Hüllermeier, E. (2018). Dyad ranking using Plackett--Luce models based on joint feature representations. Machine Learning, 107(5), 903–941. https://doi.org/10.1007/s10994-017-5694-9
traingetConfiguration, predict, predict, setConfiguration