As one of the most promising classification approaches , association rule-based classification (also called associative classification , AC) enables effectively integrating the classification tasks with association rule discovery techniques for deriving accurate, robust and interpretable results. The advantages of association rule discovery techniques over deep learning architectures in classification are mainly reflected by the aspects of better interpretability for users and higher accuracy for small sample data. Despite of great progress in both theoretical and applied aspects, there remains a lack of comprehensive and systematic overview for the recent development in AC. In light of this, this paper first conducts a statistical analysis of academic reports and related application achievements over the past decades, among which 317 are method-oriented and 200 are application-oriented. After that, through performing an in-depth analysis for these literatures, it then provides a comprehensive review for the overall learning framework, theoretical methodologies, application domains, as well as the key research challenges in the field of AC. Finally, this review displays some potential and meaningful research directions in the future by integrating the challenges with development trends, such as deep associative learning framework, human–machine intelligent associative system and semi-supervised learning within evidential framework, with the hope of assisting interested researchers gaining a quick understanding for the development trends in AC.

In this paper, a rule learning method based on interval-constrained clustering is proposed to efficiently design a compact belief rule-based classifier. The main idea of this method is to learn a compact belief rule base based on a set of prototypes generated from the original training set. First, an interval-constrained clustering algorithm is used to divide the training data for each class into several clusters, with which the number of data belonging to each cluster can be constrained within a given interval. Then, we define a belief rule based on the centroid of each cluster. Finally, a two-objective optimization procedure is designed to get a compact belief rule base with a better trade-off between accuracy and interpretability. Two experiments based on synthetic and benchmark data sets have been carried out to evaluate the performance of the proposed classifier.

In multiple attribute decision analysis (MADA) problems, one often needs to deal with assessment information with uncertainty. The evidential reasoning approach is one of the most effective methods to deal with such MADA problems. As a kernel of the evidential reasoning approach, an original evidential reasoning (ER) algorithm was firstly proposed by Yang et al, and later they modified the ER algorithm in order to satisfy the proposed four synthesis axioms. However, up to the present, the essential difference of the two ER algorithms is still unclear. In this paper, we analyze the ER algorithms in the Dempster-Shafer theory framework and prove that the original ER algorithm follows the reliability discounting and combination scheme, whereas the modified one follows the importance discounting and combination scheme. Based on these new findings, an extended ER (E2R) algorithm is proposed to take into account both the reliability and importance of different attributes, which provides a more general attribute aggregation scheme for MADA with uncertainty. A motorcycle performance assessment problem is examined to illustrate the proposed algorithm.