To date, a large number of active learning algorithms have been proposed, but active learning methods for ordinal classification are under-researched. For ordinal classification, there is a total ordering among the data classes, and it is natural that the cost of misclassifying an instance as an adjacent class should be lower than that of misclassifying it as a more disparate class. However, existing active learning algorithms typically do not consider the above ordering information in query selection. Thus, most of them do not perform satisfactorily in ordinal classification. This study proposes an active learning method for ordinal classification by considering the ordering information among classes. We design an expected cost minimization criterion that imbues the ordering information. Meanwhile, we incorporate it with an uncertainty sampling criterion to impose the query instance more informative. Furthermore, we introduce a candidate subset selection method based on the k-means algorithm to reduce the computational overhead led by the calculation of expected cost. Extensive experiments on nine public ordinal classification datasets demonstrate that the proposed method outperforms several baseline methods.
Although a plethora of research has been conducted on active learning, little research attention has been focused on active learning for ordinal classification. Traditional multi-class active learning methods usually perform unsatisfactorily in ordinal classification since they are typically designed for nominal multi-class classification. Ordinal classification is a special multi-class classification problem in which the ordering information that the cost of misclassifying an instance as an adjacent class is lower than that of misclassifying it as a more disparate class is commonly used in the solution. However, traditional active learning methods typically do not consider this ordering information during query selection. In this study, we design a novel adaptive hybrid active learning method for ordinal classification by considering the ordering information. In the proposed method, an uncertainty measure is introduced to select the hard-to-predict instances distributed between adjacent classes. To alleviate the potential sampling redundancy, we introduce a diversity measure to the uncertainty measure. Finally, we design an expected cost minimization measure by considering the ordering information to balance the contributions of the uncertainty and diversity measures and prompt the algorithm to select the instances most likely to decrease the misclassification cost of the model. Through extensive experiments on eleven public ordinal datasets, we demonstrate the superiority of the proposed method over several state-of-the-art methods.