Hierarchical sequential three-way decision model

Abstract Knowledge acquisition is one of the important issues in granular computing. In recent years, scholars have paid much attention to this problem and proposed the rule-based acquisition algorithms. However, a large number of the decision rules mined by the existing algorithms are not comprehensible. At the same time, the long detailed rules are easy to lead to over-fitting. In order to generate simpler and easier-to-comprehensible rules and improve human decision-making, a hierarchical sequential three-way decision model is proposed by combining sequential three-way decisions with hierarchical rough set model. Specifically, we generalize the concepts of the conditional attributes through the concept hierarchy tree, design the multi-hierarchical decision table with multiple levels of granularity, and illustrate the corresponding algorithm to acquire the generalized rules step by step. The experimental results demonstrate that the proposed model can mine hierarchical sequential three-way decision rules under different levels of granularity.